diff --git a/public/models.json b/public/models.json index 20c29b61e..dbdafe801 100644 --- a/public/models.json +++ b/public/models.json @@ -4,973 +4,997 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 128000, - "created": 1726099200, - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "openai/o1-mini-2024-09-12", - "name": "OpenAI: o1-mini (2024-09-12)", + "context_length": 256000, + "created": 1724371200, + "description": "Jamba 1.5 Large is part of AI21's new family of open models, offering superior speed, efficiency, and quality.\n\nIt features a 256K effective context window, the longest among open models, enabling improved performance on tasks like document summarization and analysis.\n\nBuilt on a novel SSM-Transformer architecture, it outperforms larger models like Llama 3.1 70B on benchmarks while maintaining resource efficiency.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", + "id": "ai21/jamba-1-5-large", + "name": "AI21: Jamba 1.5 Large", "per_request_limits": null, "pricing": { - "completion": "0.000012", + "completion": "0.000008", "image": "0", - "prompt": "0.000003", + "prompt": "0.000002", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 65536 + "context_length": 256000, + "is_moderated": false, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 128000, - "created": 1726099200, - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "openai/o1-mini", - "name": "OpenAI: o1-mini", + "context_length": 256000, + "created": 1724371200, + "description": "Jamba 1.5 Mini is the world's first production-grade Mamba-based model, combining SSM and Transformer architectures for a 256K context window and high efficiency.\n\nIt works with 9 languages and can handle various writing and analysis tasks as well as or better than similar small models.\n\nThis model uses less computer memory and works faster with longer texts than previous designs.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", + "id": "ai21/jamba-1-5-mini", + "name": "AI21: Jamba 1.5 Mini", "per_request_limits": null, "pricing": { - "completion": "0.000012", + "completion": "0.0000004", "image": "0", - "prompt": "0.000003", + "prompt": "0.0000002", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 65536 + "context_length": 256000, + "is_moderated": false, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 128000, - "created": 1726099200, - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "openai/o1-preview-2024-09-12", - "name": "OpenAI: o1-preview (2024-09-12)", + "context_length": 256000, + "created": 1719273600, + "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", + "id": "ai21/jamba-instruct", + "name": "AI21: Jamba Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00006", + "completion": "0.0000007", "image": "0", - "prompt": "0.000015", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 32768 + "context_length": 256000, + "is_moderated": false, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "airoboros", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama2" }, - "context_length": 128000, - "created": 1726099200, - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "openai/o1-preview", - "name": "OpenAI: o1-preview", + "context_length": 4096, + "created": 1698537600, + "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", + "id": "jondurbin/airoboros-l2-70b", + "name": "Airoboros 70B", "per_request_limits": null, "pricing": { - "completion": "0.00006", + "completion": "0.0000005", "image": "0", - "prompt": "0.000015", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 32768 + "context_length": 4096, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Llama3" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 131072, - "created": 1725580800, - "description": "Reflection Llama-3.1 70B is trained with a new technique called Reflection-Tuning that teaches a LLM to detect mistakes in its reasoning and correct course.\n\nThe model was trained on synthetic data.\n\n_These are free, rate-limited endpoints for [Reflection 70B](/models/mattshumer/reflection-70b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "mattshumer/reflection-70b:free", - "name": "Reflection 70B (free)", + "context_length": 200000, + "created": 1710288000, + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", + "id": "anthropic/claude-3-haiku", + "name": "Anthropic: Claude 3 Haiku", "per_request_limits": null, "pricing": { - "completion": "0", - "image": "0", - "prompt": "0", + "completion": "0.00000125", + "image": "0.0004", + "prompt": "0.00000025", "request": "0" }, "top_provider": { - "context_length": 8192, - "is_moderated": false, + "context_length": 200000, + "is_moderated": true, "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Llama3" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 131072, - "created": 1725580800, - "description": "Reflection Llama-3.1 70B is trained with a new technique called Reflection-Tuning that teaches a LLM to detect mistakes in its reasoning and correct course.\n\nThe model was trained on synthetic data.", - "id": "mattshumer/reflection-70b", - "name": "Reflection 70B", + "context_length": 200000, + "created": 1710288000, + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-haiku) variant._", + "id": "anthropic/claude-3-haiku:beta", + "name": "Anthropic: Claude 3 Haiku (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.0000004", - "image": "0", - "prompt": "0.00000035", + "completion": "0.00000125", + "image": "0.0004", + "prompt": "0.00000025", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Cohere" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1725062400, - "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r-03-2024", - "name": "Cohere: Command R (03-2024)", + "context_length": 200000, + "created": 1709596800, + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "id": "anthropic/claude-3-opus", + "name": "Anthropic: Claude 3 Opus", "per_request_limits": null, "pricing": { - "completion": "0.0000015", - "image": "0", - "prompt": "0.0000005", + "completion": "0.000075", + "image": "0.024", + "prompt": "0.000015", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": false, - "max_completion_tokens": 4000 + "context_length": 200000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Cohere" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1725062400, - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r-plus-04-2024", - "name": "Cohere: Command R+ (04-2024)", + "context_length": 200000, + "created": 1709596800, + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-opus) variant._", + "id": "anthropic/claude-3-opus:beta", + "name": "Anthropic: Claude 3 Opus (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0", - "prompt": "0.000003", + "completion": "0.000075", + "image": "0.024", + "prompt": "0.000015", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": 4000 + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Cohere" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1724976000, - "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r-plus-08-2024", - "name": "Cohere: Command R+ (08-2024)", + "context_length": 200000, + "created": 1709596800, + "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "id": "anthropic/claude-3-sonnet", + "name": "Anthropic: Claude 3 Sonnet", "per_request_limits": null, "pricing": { - "completion": "0.00001", - "image": "0", - "prompt": "0.0000025", + "completion": "0.000015", + "image": "0.0048", + "prompt": "0.000003", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": false, - "max_completion_tokens": 4000 + "context_length": 200000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text->text", - "tokenizer": "Cohere" + "modality": "text+image->text", + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1724976000, - "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r-08-2024", - "name": "Cohere: Command R (08-2024)", + "context_length": 200000, + "created": 1709596800, + "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-sonnet) variant._", + "id": "anthropic/claude-3-sonnet:beta", + "name": "Anthropic: Claude 3 Sonnet (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.0000006", - "image": "0", - "prompt": "0.00000015", + "completion": "0.000015", + "image": "0.0048", + "prompt": "0.000003", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": 4000 + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, "modality": "text+image->text", - "tokenizer": "Gemini" + "tokenizer": "Claude" }, - "context_length": 4000000, - "created": 1724803200, - "description": "Gemini 1.5 Flash 8B Experimental is an experimental, 8B parameter version of the [Gemini 1.5 Flash](/models/google/gemini-flash-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "google/gemini-flash-8b-1.5-exp", - "name": "Google: Gemini Flash 8B 1.5 Experimental", + "context_length": 200000, + "created": 1718841600, + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", + "id": "anthropic/claude-3.5-sonnet", + "name": "Anthropic: Claude 3.5 Sonnet", "per_request_limits": null, "pricing": { - "completion": "0", - "image": "0", - "prompt": "0", + "completion": "0.000015", + "image": "0.0048", + "prompt": "0.000003", "request": "0" }, "top_provider": { - "context_length": 4000000, - "is_moderated": false, - "max_completion_tokens": 32768 + "context_length": 200000, + "is_moderated": true, + "max_completion_tokens": 8192 } }, { "architecture": { "instruct_type": null, "modality": "text+image->text", - "tokenizer": "Gemini" + "tokenizer": "Claude" }, - "context_length": 4000000, - "created": 1724803200, - "description": "Gemini 1.5 Flash Experimental is an experimental version of the [Gemini 1.5 Flash](/models/google/gemini-flash-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "google/gemini-flash-1.5-exp", - "name": "Google: Gemini Flash 1.5 Experimental", + "context_length": 200000, + "created": 1718841600, + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3.5-sonnet) variant._", + "id": "anthropic/claude-3.5-sonnet:beta", + "name": "Anthropic: Claude 3.5 Sonnet (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0", - "image": "0", - "prompt": "0", + "completion": "0.000015", + "image": "0.0048", + "prompt": "0.000003", "request": "0" }, "top_provider": { - "context_length": 4000000, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 8192, - "created": 1724803200, - "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", - "id": "sao10k/l3.1-euryale-70b", - "name": "Llama 3.1 Euryale 70B v2.2", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", + "id": "anthropic/claude-instant-1", + "name": "Anthropic: Claude Instant v1", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0.0000024", "image": "0", - "prompt": "0.0000015", + "prompt": "0.0000008", "request": "0" }, "top_provider": { - "context_length": 16000, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 100000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Claude" }, - "context_length": 256000, - "created": 1724371200, - "description": "Jamba 1.5 Large is part of AI21's new family of open models, offering superior speed, efficiency, and quality.\n\nIt features a 256K effective context window, the longest among open models, enabling improved performance on tasks like document summarization and analysis.\n\nBuilt on a novel SSM-Transformer architecture, it outperforms larger models like Llama 3.1 70B on benchmarks while maintaining resource efficiency.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", - "id": "ai21/jamba-1-5-large", - "name": "AI21: Jamba 1.5 Large", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-instant-1) variant._", + "id": "anthropic/claude-instant-1:beta", + "name": "Anthropic: Claude Instant v1 (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.000008", + "completion": "0.0000024", "image": "0", - "prompt": "0.000002", + "prompt": "0.0000008", "request": "0" }, "top_provider": { - "context_length": 256000, + "context_length": 100000, "is_moderated": false, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "claude", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Claude" }, - "context_length": 256000, - "created": 1724371200, - "description": "Jamba 1.5 Mini is the world's first production-grade Mamba-based model, combining SSM and Transformer architectures for a 256K context window and high efficiency.\n\nIt works with 9 languages and can handle various writing and analysis tasks as well as or better than similar small models.\n\nThis model uses less computer memory and works faster with longer texts than previous designs.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", - "id": "ai21/jamba-1-5-mini", - "name": "AI21: Jamba 1.5 Mini", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", + "id": "anthropic/claude-instant-1.0", + "name": "Anthropic: Claude Instant v1.0", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0.0000024", "image": "0", - "prompt": "0.0000002", + "prompt": "0.0000008", "request": "0" }, "top_provider": { - "context_length": 256000, - "is_moderated": false, + "context_length": 100000, + "is_moderated": true, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "claude", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1724198400, - "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", - "id": "microsoft/phi-3.5-mini-128k-instruct", - "name": "Phi-3.5 Mini 128K Instruct", + "context_length": 100000, + "created": 1700611200, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", + "id": "anthropic/claude-instant-1.1", + "name": "Anthropic: Claude Instant v1.1", "per_request_limits": null, "pricing": { - "completion": "0.0000001", + "completion": "0.0000024", "image": "0", - "prompt": "0.0000001", + "prompt": "0.0000008", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 100000, + "is_moderated": true, + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "claude", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 131072, - "created": 1723939200, - "description": "Hermes 3 is a generalist language model with many improvements over [Hermes 2](/models/nousresearch/nous-hermes-2-mistral-7b-dpo), including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 70B is a competitive, if not superior finetune of the [Llama-3.1 70B foundation model](/models/meta-llama/llama-3.1-70b-instruct), focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.", - "id": "nousresearch/hermes-3-llama-3.1-70b", - "name": "Nous: Hermes 3 70B Instruct", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", + "id": "anthropic/claude-1", + "name": "Anthropic: Claude v1", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0.000024", "image": "0", - "prompt": "0.0000004", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 12288, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 100000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "claude", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 131072, - "created": 1723766400, - "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.", - "id": "nousresearch/hermes-3-llama-3.1-405b", - "name": "Nous: Hermes 3 405B Instruct", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", + "id": "anthropic/claude-1.2", + "name": "Anthropic: Claude v1.2", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.000024", "image": "0", - "prompt": "0", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 18000, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 100000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 131072, - "created": 1723766400, - "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.\n\n_These are free, rate-limited endpoints for [Hermes 3 405B Instruct](/models/nousresearch/hermes-3-llama-3.1-405b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "nousresearch/hermes-3-llama-3.1-405b:free", - "name": "Nous: Hermes 3 405B Instruct (free)", + "context_length": 200000, + "created": 1700611200, + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "id": "anthropic/claude-2", + "name": "Anthropic: Claude v2", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.000024", "image": "0", - "prompt": "0", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 8192, - "is_moderated": false, + "context_length": 200000, + "is_moderated": true, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1723766400, - "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.\n\n_These are extended-context endpoints for [Hermes 3 405B Instruct](/models/nousresearch/hermes-3-llama-3.1-405b). They may have higher prices._", - "id": "nousresearch/hermes-3-llama-3.1-405b:extended", - "name": "Nous: Hermes 3 405B Instruct (extended)", + "context_length": 200000, + "created": 1700611200, + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2) variant._", + "id": "anthropic/claude-2:beta", + "name": "Anthropic: Claude v2 (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.000024", "image": "0", - "prompt": "0", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 127072, - "created": 1723593600, - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance. The model is built upon the Llama 3.1 405B and has internet access.", - "id": "perplexity/llama-3.1-sonar-huge-128k-online", - "name": "Perplexity: Llama 3.1 Sonar 405B Online", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", + "id": "anthropic/claude-2.0", + "name": "Anthropic: Claude v2.0", "per_request_limits": null, "pricing": { - "completion": "0.000005", + "completion": "0.000024", "image": "0", - "prompt": "0.000005", - "request": "0.005" + "prompt": "0.000008", + "request": "0" }, "top_provider": { - "context_length": 127072, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 100000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "modality": "text->text", + "tokenizer": "Claude" }, - "context_length": 128000, - "created": 1723593600, - "description": "Dynamic model continuously updated to the current version of [GPT-4o](/models/openai/gpt-4o) in ChatGPT. Intended for research and evaluation.\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "openai/chatgpt-4o-latest", - "name": "OpenAI: ChatGPT-4o", + "context_length": 100000, + "created": 1690502400, + "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2.0) variant._", + "id": "anthropic/claude-2.0:beta", + "name": "Anthropic: Claude v2.0 (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.007225", - "prompt": "0.000005", + "completion": "0.000024", + "image": "0", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 16384 + "context_length": 100000, + "is_moderated": false, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Claude" }, - "context_length": 8192, - "created": 1723507200, - "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", - "id": "sao10k/l3-lunaris-8b", - "name": "Llama 3 8B Lunaris", - "per_request_limits": null, - "pricing": { - "completion": "0.000002", - "image": "0", - "prompt": "0.000002", + "context_length": 200000, + "created": 1700611200, + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "id": "anthropic/claude-2.1", + "name": "Anthropic: Claude v2.1", + "per_request_limits": null, + "pricing": { + "completion": "0.000024", + "image": "0", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 8192, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 200000, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Claude" }, - "context_length": 12000, - "created": 1723507200, - "description": "Starcannon 12B is a creative roleplay and story writing model, using [nothingiisreal/mn-celeste-12b](https://openrouter.ai/models/nothingiisreal/mn-celeste-12b) as a base and [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", - "id": "aetherwiing/mn-starcannon-12b", - "name": "Mistral Nemo 12B Starcannon", + "context_length": 200000, + "created": 1700611200, + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2.1) variant._", + "id": "anthropic/claude-2.1:beta", + "name": "Anthropic: Claude v2.1 (self-moderated)", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.000024", "image": "0", - "prompt": "0.000002", + "prompt": "0.000008", "request": "0" }, "top_provider": { - "context_length": 12000, + "context_length": 200000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "modality": "text->text", + "tokenizer": "Router" }, - "context_length": 128000, - "created": 1722902400, - "description": "The 2024-08-06 version of GPT-4o offers improved performance in structured outputs, with the ability to supply a JSON schema in the respone_format. Read more [here](https://openai.com/index/introducing-structured-outputs-in-the-api/).\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", - "id": "openai/gpt-4o-2024-08-06", - "name": "OpenAI: GPT-4o (2024-08-06)", + "context_length": 200000, + "created": 1699401600, + "description": "Depending on their size, subject, and complexity, your prompts will be sent to [Llama 3 70B Instruct](/models/meta-llama/llama-3-70b-instruct), [Claude 3.5 Sonnet (self-moderated)](/models/anthropic/claude-3.5-sonnet:beta) or [GPT-4o](/models/openai/gpt-4o). To see which model was used, visit [Activity](/activity).\n\nA major redesign of this router is coming soon. Stay tuned on [Discord](https://discord.gg/fVyRaUDgxW) for updates.", + "id": "openrouter/auto", + "name": "Auto (best for prompt)", "per_request_limits": null, "pricing": { - "completion": "0.00001", - "image": "0.0036125", - "prompt": "0.0000025", - "request": "0" + "completion": "-1", + "image": "-1", + "prompt": "-1", + "request": "-1" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 16384 + "context_length": null, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "none", + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Llama2" }, - "context_length": 131072, - "created": 1722556800, - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3.1-405b", - "name": "Meta: Llama 3.1 405B (base)", + "context_length": 4096, + "created": 1704412800, + "description": "A 75/25 merge of [Chronos 13b v2](https://huggingface.co/elinas/chronos-13b-v2) and [Nous Hermes Llama2 13b](/models/nousresearch/nous-hermes-llama2-13b). This offers the imaginative writing style of Chronos while retaining coherency. Outputs are long and use exceptional prose. #merge", + "id": "austism/chronos-hermes-13b", + "name": "Chronos Hermes 13B v2", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.00000013", "image": "0", - "prompt": "0.000002", + "prompt": "0.00000013", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Cohere" }, - "context_length": 32000, - "created": 1722556800, - "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", - "id": "nothingiisreal/mn-celeste-12b", - "name": "Mistral Nemo 12B Celeste", + "context_length": 4096, + "created": 1710374400, + "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command", + "name": "Cohere: Command", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0.0000019", "image": "0", - "prompt": "0.0000015", + "prompt": "0.00000095", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Gemini" + "modality": "text->text", + "tokenizer": "Cohere" }, - "context_length": 4000000, - "created": 1722470400, - "description": "Gemini 1.5 Pro (0827) is an experimental version of the [Gemini 1.5 Pro](/models/google/gemini-pro-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "id": "google/gemini-pro-1.5-exp", - "name": "Google: Gemini Pro 1.5 Experimental", + "context_length": 128000, + "created": 1710374400, + "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r", + "name": "Cohere: Command R", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.000001425", "image": "0", - "prompt": "0", + "prompt": "0.000000475", "request": "0" }, "top_provider": { - "context_length": 4000000, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": 4000 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Cohere" }, - "context_length": 127072, - "created": 1722470400, - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "id": "perplexity/llama-3.1-sonar-large-128k-online", - "name": "Perplexity: Llama 3.1 Sonar 70B Online", + "context_length": 128000, + "created": 1709341200, + "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r-03-2024", + "name": "Cohere: Command R (03-2024)", "per_request_limits": null, "pricing": { - "completion": "0.000001", + "completion": "0.000001425", "image": "0", - "prompt": "0.000001", - "request": "0.005" + "prompt": "0.000000475", + "request": "0" }, "top_provider": { - "context_length": 127072, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Cohere" }, - "context_length": 131072, - "created": 1722470400, - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3.1-sonar-large-128k-online) of this model has Internet access.", - "id": "perplexity/llama-3.1-sonar-large-128k-chat", - "name": "Perplexity: Llama 3.1 Sonar 70B", + "context_length": 128000, + "created": 1724976000, + "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r-08-2024", + "name": "Cohere: Command R (08-2024)", "per_request_limits": null, "pricing": { - "completion": "0.000001", + "completion": "0.00000057", "image": "0", - "prompt": "0.000001", + "prompt": "0.0000001425", "request": "0" }, "top_provider": { - "context_length": 131072, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Cohere" }, - "context_length": 127072, - "created": 1722470400, - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "id": "perplexity/llama-3.1-sonar-small-128k-online", - "name": "Perplexity: Llama 3.1 Sonar 8B Online", + "context_length": 128000, + "created": 1712188800, + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r-plus", + "name": "Cohere: Command R+", "per_request_limits": null, "pricing": { - "completion": "0.0000002", + "completion": "0.00001425", "image": "0", - "prompt": "0.0000002", - "request": "0.005" + "prompt": "0.00000285", + "request": "0" }, "top_provider": { - "context_length": 127072, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Cohere" }, - "context_length": 131072, - "created": 1722470400, - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3.1-sonar-small-128k-online) of this model has Internet access.", - "id": "perplexity/llama-3.1-sonar-small-128k-chat", - "name": "Perplexity: Llama 3.1 Sonar 8B", + "context_length": 128000, + "created": 1712016000, + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r-plus-04-2024", + "name": "Cohere: Command R+ (04-2024)", "per_request_limits": null, "pricing": { - "completion": "0.0000002", + "completion": "0.00001425", "image": "0", - "prompt": "0.0000002", + "prompt": "0.00000285", "request": "0" }, "top_provider": { - "context_length": 131072, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Cohere" }, - "context_length": 131072, - "created": 1721692800, - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 70B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3.1-70b-instruct", - "name": "Meta: Llama 3.1 70B Instruct", + "context_length": 128000, + "created": 1724976000, + "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", + "id": "cohere/command-r-plus-08-2024", + "name": "Cohere: Command R+ (08-2024)", "per_request_limits": null, "pricing": { - "completion": "0.0000003", + "completion": "0.0000095", "image": "0", - "prompt": "0.0000003", + "prompt": "0.000002375", "request": "0" }, "top_provider": { - "context_length": 131072, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4000 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Other" }, - "context_length": 131072, - "created": 1721692800, - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.1 8B Instruct](/models/meta-llama/llama-3.1-8b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "meta-llama/llama-3.1-8b-instruct:free", - "name": "Meta: Llama 3.1 8B Instruct (free)", + "context_length": 32768, + "created": 1711670400, + "description": "DBRX is a new open source large language model developed by Databricks. At 132B, it outperforms existing open source LLMs like Llama 2 70B and [Mixtral-8x7b](/models/mistralai/mixtral-8x7b) on standard industry benchmarks for language understanding, programming, math, and logic.\n\nIt uses a fine-grained mixture-of-experts (MoE) architecture. 36B parameters are active on any input. It was pre-trained on 12T tokens of text and code data. Compared to other open MoE models like Mixtral-8x7B and Grok-1, DBRX is fine-grained, meaning it uses a larger number of smaller experts.\n\nSee the launch announcement and benchmark results [here](https://www.databricks.com/blog/introducing-dbrx-new-state-art-open-llm).\n\n#moe", + "id": "databricks/dbrx-instruct", + "name": "Databricks: DBRX 132B Instruct", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.00000108", "image": "0", - "prompt": "0", + "prompt": "0.00000108", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Other" }, - "context_length": 131072, - "created": 1721692800, - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3.1-8b-instruct", - "name": "Meta: Llama 3.1 8B Instruct", + "context_length": 128000, + "created": 1715644800, + "description": "DeepSeek-V2.5 is an upgraded version that combines DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct. The new model integrates the general and coding abilities of the two previous versions.\n\nDeepSeek-V2 Chat is a conversational finetune of DeepSeek-V2, a Mixture-of-Experts (MoE) language model. It comprises 236B total parameters, of which 21B are activated for each token.\n\nCompared with DeepSeek 67B, DeepSeek-V2 achieves stronger performance, and meanwhile saves 42.5% of training costs, reduces the KV cache by 93.3%, and boosts the maximum generation throughput to 5.76 times.\n\nDeepSeek-V2 achieves remarkable performance on both standard benchmarks and open-ended generation evaluations.", + "id": "deepseek/deepseek-chat", + "name": "DeepSeek V2.5", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.00000028", "image": "0", - "prompt": "0.000000055", + "prompt": "0.00000014", "request": "0" }, "top_provider": { - "context_length": 100000, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Mistral" }, - "context_length": 131072, - "created": 1721692800, - "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3.1-405b-instruct", - "name": "Meta: Llama 3.1 405B Instruct", + "context_length": 32768, + "created": 1703116800, + "description": "This is a 16k context fine-tune of [Mixtral-8x7b](/models/mistralai/mixtral-8x7b). It excels in coding tasks due to extensive training with coding data and is known for its obedience, although it lacks DPO tuning.\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored", + "id": "cognitivecomputations/dolphin-mixtral-8x7b", + "name": "Dolphin 2.6 Mixtral 8x7B 🐬", "per_request_limits": null, "pricing": { - "completion": "0.00000179", + "completion": "0.0000005", "image": "0", - "prompt": "0.00000179", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "chatml", "modality": "text->text", "tokenizer": "Mistral" }, - "context_length": 256000, - "created": 1721347200, - "description": "A 7.3B parameter Mamba-based model designed for code and reasoning tasks.\n\n- Linear time inference, allowing for theoretically infinite sequence lengths\n- 256k token context window\n- Optimized for quick responses, especially beneficial for code productivity\n- Performs comparably to state-of-the-art transformer models in code and reasoning tasks\n- Available under the Apache 2.0 license for free use, modification, and distribution", - "id": "mistralai/codestral-mamba", - "name": "Mistral: Codestral Mamba", + "context_length": 65536, + "created": 1717804800, + "description": "Dolphin 2.9 is designed for instruction following, conversational, and coding. This model is a finetune of [Mixtral 8x22B Instruct](/models/mistralai/mixtral-8x22b-instruct). It features a 64k context length and was fine-tuned with a 16k sequence length using ChatML templates.\n\nThis model is a successor to [Dolphin Mixtral 8x7B](/models/cognitivecomputations/dolphin-mixtral-8x7b).\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored", + "id": "cognitivecomputations/dolphin-mixtral-8x22b", + "name": "Dolphin 2.9.2 Mixtral 8x22B 🐬", "per_request_limits": null, "pricing": { - "completion": "0.00000025", + "completion": "0.0000009", "image": "0", - "prompt": "0.00000025", + "prompt": "0.0000009", "request": "0" }, "top_provider": { - "context_length": 256000, + "context_length": 16000, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Qwen" }, - "context_length": 128000, - "created": 1721347200, - "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", - "id": "mistralai/mistral-nemo", - "name": "Mistral: Mistral Nemo", + "context_length": 32768, + "created": 1727654400, + "description": "A model specializing in RP and creative writing, this model is based on Qwen2.5-14B, fine-tuned with a mixture of synthetic and natural data.\n\nIt is trained on 1.5M tokens of role-play data, and fine-tuned on 1.5M tokens of synthetic data.", + "id": "eva-unit-01/eva-qwen-2.5-14b", + "name": "EVA Qwen2.5 14B", "per_request_limits": null, "pricing": { - "completion": "0.00000013", + "completion": "0.0000005", "image": "0", - "prompt": "0.00000013", + "prompt": "0.00000025", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "instruct_type": "alpaca", + "modality": "text->text", + "tokenizer": "Llama2" }, - "context_length": 128000, - "created": 1721260800, - "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.", - "id": "openai/gpt-4o-mini-2024-07-18", - "name": "OpenAI: GPT-4o-mini (2024-07-18)", + "context_length": 8192, + "created": 1713657600, + "description": "Creative writing model, routed with permission. It's fast, it keeps the conversation going, and it stays in character.\n\nIf you submit a raw prompt, you can use Alpaca or Vicuna formats.", + "id": "sao10k/fimbulvetr-11b-v2", + "name": "Fimbulvetr 11B v2", "per_request_limits": null, "pricing": { - "completion": "0.0000006", - "image": "0.007225", - "prompt": "0.00000015", + "completion": "0.0000015", + "image": "0", + "prompt": "0.000000375", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 16384 + "context_length": 8192, + "is_moderated": false, + "max_completion_tokens": 2048 + } + }, + { + "architecture": { + "instruct_type": "airoboros", + "modality": "text->text", + "tokenizer": "Llama2" + }, + "context_length": 6144, + "created": 1699574400, + "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", + "id": "alpindale/goliath-120b", + "name": "Goliath 120B", + "per_request_limits": null, + "pricing": { + "completion": "0.000009375", + "image": "0", + "prompt": "0.000009375", + "request": "0" + }, + "top_provider": { + "context_length": 6144, + "is_moderated": false, + "max_completion_tokens": 400 } }, { "architecture": { "instruct_type": null, "modality": "text+image->text", - "tokenizer": "GPT" + "tokenizer": "Gemini" }, - "context_length": 128000, - "created": 1721260800, - "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.", - "id": "openai/gpt-4o-mini", - "name": "OpenAI: GPT-4o-mini", + "context_length": 1048576, + "created": 1715644800, + "description": "Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.\n\nGemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", + "id": "google/gemini-flash-1.5", + "name": "Google: Gemini Flash 1.5", "per_request_limits": null, "pricing": { - "completion": "0.0000006", - "image": "0.007225", - "prompt": "0.00000015", + "completion": "0.0000003", + "image": "0.00004", + "prompt": "0.000000075", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 16384 + "context_length": 1048576, + "is_moderated": false, + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "chatml", - "modality": "text->text", - "tokenizer": "Qwen" + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "Gemini" }, - "context_length": 32768, - "created": 1721088000, - "description": "Qwen2 7B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).\n\n_These are free, rate-limited endpoints for [Qwen 2 7B Instruct](/models/qwen/qwen-2-7b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "qwen/qwen-2-7b-instruct:free", - "name": "Qwen 2 7B Instruct (free)", + "context_length": 1048576, + "created": 1724803200, + "description": "Gemini 1.5 Flash Experimental is an experimental version of the [Gemini 1.5 Flash](/models/google/gemini-flash-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "google/gemini-flash-1.5-exp", + "name": "Google: Gemini Flash 1.5 Experimental", "per_request_limits": null, "pricing": { "completion": "0", @@ -979,127 +1003,127 @@ "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 1048576, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "chatml", - "modality": "text->text", - "tokenizer": "Qwen" + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "Gemini" }, - "context_length": 32768, - "created": 1721088000, - "description": "Qwen2 7B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "id": "qwen/qwen-2-7b-instruct", - "name": "Qwen 2 7B Instruct", + "context_length": 1048576, + "created": 1724803200, + "description": "Gemini 1.5 Flash 8B Experimental is an experimental, 8B parameter version of the [Gemini 1.5 Flash](/models/google/gemini-flash-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "google/gemini-flash-8b-1.5-exp", + "name": "Google: Gemini Flash 8B 1.5 Experimental", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0", "image": "0", - "prompt": "0.000000055", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 1048576, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "gemma", + "instruct_type": null, "modality": "text->text", "tokenizer": "Gemini" }, - "context_length": 8192, - "created": 1720828800, - "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", - "id": "google/gemma-2-27b-it", - "name": "Google: Gemma 2 27B", + "context_length": 32760, + "created": 1702425600, + "description": "Google's flagship text generation model. Designed to handle natural language tasks, multiturn text and code chat, and code generation.\n\nSee the benchmarks and prompting guidelines from [Deepmind](https://deepmind.google/technologies/gemini/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", + "id": "google/gemini-pro", + "name": "Google: Gemini Pro 1.0", "per_request_limits": null, "pricing": { - "completion": "0.00000027", - "image": "0", - "prompt": "0.00000027", + "completion": "0.0000015", + "image": "0.0025", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 32760, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "chatml", - "modality": "text->text", - "tokenizer": "Qwen" + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "Gemini" }, - "context_length": 16384, - "created": 1720656000, - "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", - "id": "alpindale/magnum-72b", - "name": "Magnum 72B", + "context_length": 2097152, + "created": 1712620800, + "description": "Google's latest multimodal model, supporting image and video in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", + "id": "google/gemini-pro-1.5", + "name": "Google: Gemini Pro 1.5", "per_request_limits": null, "pricing": { - "completion": "0.0000045", - "image": "0", - "prompt": "0.00000375", + "completion": "0.0000105", + "image": "0.00263", + "prompt": "0.0000035", "request": "0" }, "top_provider": { - "context_length": 16384, + "context_length": 2097152, "is_moderated": false, - "max_completion_tokens": 1024 + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "chatml", - "modality": "text->text", - "tokenizer": "Llama3" + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "Gemini" }, - "context_length": 16384, - "created": 1720656000, - "description": "An experimental merge model based on Llama 3, exhibiting a very distinctive style of writing. It combines the the best of [Meta's Llama 3 8B](https://openrouter.ai/models/meta-llama/llama-3-8b-instruct) and Nous Research's [Hermes 2 Pro](https://openrouter.ai/models/nousresearch/hermes-2-pro-llama-3-8b).\n\nHermes-2 Θ (theta) was specifically designed with a few capabilities in mind: executing function calls, generating JSON output, and most remarkably, demonstrating metacognitive abilities (contemplating the nature of thought and recognizing the diversity of cognitive processes among individuals).", - "id": "nousresearch/hermes-2-theta-llama-3-8b", - "name": "Nous: Hermes 2 Theta 8B", + "context_length": 2097152, + "created": 1722470400, + "description": "Gemini 1.5 Pro (0827) is an experimental version of the [Gemini 1.5 Pro](/models/google/gemini-pro-1.5) model.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "google/gemini-pro-1.5-exp", + "name": "Google: Gemini Pro 1.5 Experimental", "per_request_limits": null, "pricing": { - "completion": "0.000001125", + "completion": "0", "image": "0", - "prompt": "0.0000001875", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 16384, + "context_length": 2097152, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "gemma", - "modality": "text->text", + "instruct_type": null, + "modality": "text+image->text", "tokenizer": "Gemini" }, - "context_length": 8192, - "created": 1719532800, - "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).\n\n_These are free, rate-limited endpoints for [Gemma 2 9B](/models/google/gemma-2-9b-it). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "google/gemma-2-9b-it:free", - "name": "Google: Gemma 2 9B (free)", + "context_length": 16384, + "created": 1702425600, + "description": "Google's flagship multimodal model, supporting image and video in text or chat prompts for a text or code response.\n\nSee the benchmarks and prompting guidelines from [Deepmind](https://deepmind.google/technologies/gemini/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", + "id": "google/gemini-pro-vision", + "name": "Google: Gemini Pro Vision 1.0", "per_request_limits": null, "pricing": { - "completion": "0", - "image": "0", - "prompt": "0", + "completion": "0.0000015", + "image": "0.0025", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 16384, "is_moderated": false, "max_completion_tokens": 2048 } @@ -1111,7 +1135,31 @@ "tokenizer": "Gemini" }, "context_length": 8192, - "created": 1719532800, + "created": 1720828800, + "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "id": "google/gemma-2-27b-it", + "name": "Google: Gemma 2 27B", + "per_request_limits": null, + "pricing": { + "completion": "0.00000027", + "image": "0", + "prompt": "0.00000027", + "request": "0" + }, + "top_provider": { + "context_length": 8192, + "is_moderated": false, + "max_completion_tokens": null + } + }, + { + "architecture": { + "instruct_type": "gemma", + "modality": "text->text", + "tokenizer": "Gemini" + }, + "context_length": 8192, + "created": 1719532800, "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", "id": "google/gemma-2-9b-it", "name": "Google: Gemma 2 9B", @@ -1130,212 +1178,212 @@ }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": "gemma", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Gemini" }, - "context_length": 32000, - "created": 1719446400, - "description": "Stheno 8B 32K is a creative writing/roleplay model from [Sao10k](https://ko-fi.com/sao10k). It was trained at 8K context, then expanded to 32K context.\n\nCompared to older Stheno version, this model is trained on:\n- 2x the amount of creative writing samples\n- Cleaned up roleplaying samples\n- Fewer low quality samples", - "id": "sao10k/l3-stheno-8b", - "name": "Llama 3 Stheno 8B v3.3 32K", + "context_length": 8192, + "created": 1719532800, + "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).\n\n_These are free, rate-limited endpoints for [Gemma 2 9B](/models/google/gemma-2-9b-it). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "google/gemma-2-9b-it:free", + "name": "Google: Gemma 2 9B (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0", "image": "0", - "prompt": "0.00000025", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "PaLM" }, - "context_length": 256000, - "created": 1719273600, - "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", - "id": "ai21/jamba-instruct", - "name": "AI21: Jamba Instruct", + "context_length": 9216, + "created": 1689811200, + "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", + "id": "google/palm-2-chat-bison", + "name": "Google: PaLM 2 Chat", "per_request_limits": null, "pricing": { - "completion": "0.0000007", + "completion": "0.000002", "image": "0", - "prompt": "0.0000005", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 256000, + "context_length": 9216, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": 1024 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "modality": "text->text", + "tokenizer": "PaLM" }, - "context_length": 200000, - "created": 1718841600, - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", - "id": "anthropic/claude-3.5-sonnet", - "name": "Anthropic: Claude 3.5 Sonnet", + "context_length": 32760, + "created": 1698969600, + "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", + "id": "google/palm-2-chat-bison-32k", + "name": "Google: PaLM 2 Chat 32k", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.0048", - "prompt": "0.000003", + "completion": "0.000002", + "image": "0", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": true, + "context_length": 32768, + "is_moderated": false, "max_completion_tokens": 8192 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "modality": "text->text", + "tokenizer": "PaLM" }, - "context_length": 200000, - "created": 1718841600, - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3.5-sonnet) variant._", - "id": "anthropic/claude-3.5-sonnet:beta", - "name": "Anthropic: Claude 3.5 Sonnet (self-moderated)", + "context_length": 7168, + "created": 1689811200, + "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", + "id": "google/palm-2-codechat-bison", + "name": "Google: PaLM 2 Code Chat", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.0048", - "prompt": "0.000003", + "completion": "0.000002", + "image": "0", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 7168, "is_moderated": false, - "max_completion_tokens": 8192 + "max_completion_tokens": 1024 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "PaLM" }, - "context_length": 8192, - "created": 1718668800, - "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", - "id": "sao10k/l3-euryale-70b", - "name": "Llama 3 Euryale 70B v2.1", + "context_length": 32760, + "created": 1698969600, + "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", + "id": "google/palm-2-codechat-bison-32k", + "name": "Google: PaLM 2 Code Chat 32k", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0.000002", "image": "0", - "prompt": "0.00000035", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 8192 } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "zephyr", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Mistral" }, - "context_length": 4000, - "created": 1718409600, - "description": "Phi-3 4K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 128k context length, try [Phi-3 Medium 128K](/models/microsoft/phi-3-medium-128k-instruct).", - "id": "microsoft/phi-3-medium-4k-instruct", - "name": "Phi-3 Medium 4K Instruct", + "context_length": 4096, + "created": 1690934400, + "description": "Zephyr is a series of language models that are trained to act as helpful assistants. Zephyr-7B-β is the second model in the series, and is a fine-tuned version of [mistralai/Mistral-7B-v0.1](/models/mistralai/mistral-7b-instruct-v0.1) that was trained on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO).\n\n_These are free, rate-limited endpoints for [Zephyr 7B](/models/huggingfaceh4/zephyr-7b-beta). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "huggingfaceh4/zephyr-7b-beta:free", + "name": "Hugging Face: Zephyr 7B (free)", "per_request_limits": null, "pricing": { - "completion": "0.00000014", + "completion": "0", "image": "0", - "prompt": "0.00000014", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 4000, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "vicuna", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Other" }, - "context_length": 65536, - "created": 1717804800, - "description": "Dolphin 2.9 is designed for instruction following, conversational, and coding. This model is a finetune of [Mixtral 8x22B Instruct](/models/mistralai/mixtral-8x22b-instruct). It features a 64k context length and was fine-tuned with a 16k sequence length using ChatML templates.\n\nThis model is a successor to [Dolphin Mixtral 8x7B](/models/cognitivecomputations/dolphin-mixtral-8x7b).\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored", - "id": "cognitivecomputations/dolphin-mixtral-8x22b", - "name": "Dolphin 2.9.2 Mixtral 8x22B 🐬", + "context_length": 32768, + "created": 1727654400, + "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.", + "id": "liquid/lfm-40b", + "name": "Liquid: LFM 40B MoE", "per_request_limits": null, "pricing": { - "completion": "0.0000009", + "completion": "0", "image": "0", - "prompt": "0.0000009", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 16000, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "vicuna", "modality": "text->text", - "tokenizer": "Qwen" + "tokenizer": "Other" }, "context_length": 32768, - "created": 1717718400, - "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "id": "qwen/qwen-2-72b-instruct", - "name": "Qwen 2 72B Instruct", + "created": 1727654400, + "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.\n\n_These are free, rate-limited endpoints for [LFM 40B MoE](/models/liquid/lfm-40b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "liquid/lfm-40b:free", + "name": "Liquid: LFM 40B MoE (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0", "image": "0", - "prompt": "0.00000035", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, "context_length": 8192, - "created": 1716768000, - "description": "Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.", - "id": "nousresearch/hermes-2-pro-llama-3-8b", - "name": "NousResearch: Hermes 2 Pro - Llama-3 8B", + "created": 1723507200, + "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", + "id": "sao10k/l3-lunaris-8b", + "name": "Llama 3 8B Lunaris", "per_request_limits": null, "pricing": { - "completion": "0.00000014", + "completion": "0.000002", "image": "0", - "prompt": "0.00000014", + "prompt": "0.000002", "request": "0" }, "top_provider": { @@ -1346,380 +1394,380 @@ }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1716768000, - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct v0.2](/models/mistralai/mistral-7b-instruct-v0.2), with the following changes:\n\n- Extended vocabulary to 32768\n- Supports v3 Tokenizer\n- Supports function calling\n\nNOTE: Support for function calling depends on the provider.", - "id": "mistralai/mistral-7b-instruct-v0.3", - "name": "Mistral: Mistral 7B Instruct v0.3", + "context_length": 8192, + "created": 1718668800, + "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", + "id": "sao10k/l3-euryale-70b", + "name": "Llama 3 Euryale 70B v2.1", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.0000004", "image": "0", - "prompt": "0.000000055", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1716768000, - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*\n\n_These are free, rate-limited endpoints for [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "mistralai/mistral-7b-instruct:free", - "name": "Mistral: Mistral 7B Instruct (free)", + "context_length": 8192, + "created": 1715817600, + "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "id": "neversleep/llama-3-lumimaid-70b", + "name": "Llama 3 Lumimaid 70B", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.0000045", "image": "0", - "prompt": "0", + "prompt": "0.000003375", "request": "0" }, "top_provider": { "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1716768000, - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", - "id": "mistralai/mistral-7b-instruct", - "name": "Mistral: Mistral 7B Instruct", + "context_length": 24576, + "created": 1714780800, + "description": "The NeverSleep team is back, with a Llama 3 8B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "id": "neversleep/llama-3-lumimaid-8b", + "name": "Llama 3 Lumimaid 8B", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.000001125", "image": "0", - "prompt": "0.000000055", + "prompt": "0.0000001875", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1716768000, - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*\n\n_These are higher-throughput endpoints for [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct). They may have higher prices._", - "id": "mistralai/mistral-7b-instruct:nitro", - "name": "Mistral: Mistral 7B Instruct (nitro)", + "context_length": 24576, + "created": 1714780800, + "description": "The NeverSleep team is back, with a Llama 3 8B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are extended-context endpoints for [Llama 3 Lumimaid v0.1 8B](/models/neversleep/llama-3-lumimaid-8b). They may have higher prices._", + "id": "neversleep/llama-3-lumimaid-8b:extended", + "name": "Llama 3 Lumimaid 8B (extended)", "per_request_limits": null, "pricing": { - "completion": "0.00000007", + "completion": "0.000001125", "image": "0", - "prompt": "0.00000007", + "prompt": "0.0000001875", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 24576, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1716681600, - "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.\n\n_These are free, rate-limited endpoints for [Phi-3 Mini 128K Instruct](/models/microsoft/phi-3-mini-128k-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "microsoft/phi-3-mini-128k-instruct:free", - "name": "Phi-3 Mini 128K Instruct (free)", + "context_length": 8192, + "created": 1724803200, + "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", + "id": "sao10k/l3.1-euryale-70b", + "name": "Llama 3.1 Euryale 70B v2.2", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.0000004", "image": "0", - "prompt": "0", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1716681600, - "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", - "id": "microsoft/phi-3-mini-128k-instruct", - "name": "Phi-3 Mini 128K Instruct", + "context_length": 131072, + "created": 1726358400, + "description": "Lumimaid v0.2 8B is a finetune of [Llama 3.1 8B](/models/meta-llama/llama-3.1-8b-instruct) with a \"HUGE step up dataset wise\" compared to Lumimaid v0.1. Sloppy chats output were purged.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "id": "neversleep/llama-3.1-lumimaid-8b", + "name": "Lumimaid v0.2 8B", "per_request_limits": null, "pricing": { - "completion": "0.0000001", + "completion": "0.000001125", "image": "0", - "prompt": "0.0000001", + "prompt": "0.0000001875", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "airoboros", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Llama2" }, - "context_length": 128000, - "created": 1716508800, - "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).\n\n_These are free, rate-limited endpoints for [Phi-3 Medium 128K Instruct](/models/microsoft/phi-3-medium-128k-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "microsoft/phi-3-medium-128k-instruct:free", - "name": "Phi-3 Medium 128K Instruct (free)", + "context_length": 4096, + "created": 1699747200, + "description": "A Mythomax/MLewd_13B-style merge of selected 70B models.\nA multi-model merge of several LLaMA2 70B finetunes for roleplaying and creative work. The goal was to create a model that combines creativity with intelligence for an enhanced experience.\n\n#merge #uncensored", + "id": "lizpreciatior/lzlv-70b-fp16-hf", + "name": "lzlv 70B", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.0000004", "image": "0", - "prompt": "0", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "phi3", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Qwen" }, - "context_length": 128000, - "created": 1716508800, - "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", - "id": "microsoft/phi-3-medium-128k-instruct", - "name": "Phi-3 Medium 128K Instruct", + "context_length": 16384, + "created": 1720656000, + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "id": "alpindale/magnum-72b", + "name": "Magnum 72B", "per_request_limits": null, "pricing": { - "completion": "0.000001", + "completion": "0.0000045", "image": "0", - "prompt": "0.000001", + "prompt": "0.00000375", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 16384, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 1024 } }, { "architecture": { - "instruct_type": "llama3", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Llama3" + "tokenizer": "Qwen" }, - "context_length": 8192, - "created": 1715817600, - "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "neversleep/llama-3-lumimaid-70b", - "name": "Llama 3 Lumimaid 70B", + "context_length": 32768, + "created": 1727654400, + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "id": "anthracite-org/magnum-v2-72b", + "name": "Magnum v2 72B", "per_request_limits": null, "pricing": { "completion": "0.0000045", "image": "0", - "prompt": "0.000003375", + "prompt": "0.00000375", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Gemini" + "instruct_type": "alpaca", + "modality": "text->text", + "tokenizer": "Llama2" }, - "context_length": 4000000, - "created": 1715644800, - "description": "Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.\n\nGemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", - "id": "google/gemini-flash-1.5", - "name": "Google: Gemini Flash 1.5", + "context_length": 8000, + "created": 1690934400, + "description": "An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.", + "id": "mancer/weaver", + "name": "Mancer: Weaver (alpha)", "per_request_limits": null, "pricing": { - "completion": "0.00000015", - "image": "0.00004", - "prompt": "0.0000000375", + "completion": "0.00000225", + "image": "0", + "prompt": "0.000001875", "request": "0" }, "top_provider": { - "context_length": 4000000, + "context_length": 8000, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": 1000 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1715644800, - "description": "DeepSeek-V2.5 is an upgraded version that combines DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct. The new model integrates the general and coding abilities of the two previous versions.\n\nDeepSeek-V2 Chat is a conversational finetune of DeepSeek-V2, a Mixture-of-Experts (MoE) language model. It comprises 236B total parameters, of which 21B are activated for each token.\n\nCompared with DeepSeek 67B, DeepSeek-V2 achieves stronger performance, and meanwhile saves 42.5% of training costs, reduces the KV cache by 93.3%, and boosts the maximum generation throughput to 5.76 times.\n\nDeepSeek-V2 achieves remarkable performance on both standard benchmarks and open-ended generation evaluations.", - "id": "deepseek/deepseek-chat", - "name": "DeepSeek V2.5", + "context_length": 8192, + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3-70b-instruct", + "name": "Meta: Llama 3 70B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00000028", + "completion": "0.0000004", "image": "0", - "prompt": "0.00000014", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 28000, - "created": 1715644800, - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-large-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "id": "perplexity/llama-3-sonar-large-32k-online", - "name": "Perplexity: Llama3 Sonar 70B Online", + "context_length": 8192, + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are higher-throughput endpoints for [Llama 3 70B Instruct](/models/meta-llama/llama-3-70b-instruct). They may have higher prices._", + "id": "meta-llama/llama-3-70b-instruct:nitro", + "name": "Meta: Llama 3 70B Instruct (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.000001", + "completion": "0.000000792", "image": "0", - "prompt": "0.000001", - "request": "0.005" + "prompt": "0.000000792", + "request": "0" }, "top_provider": { - "context_length": 28000, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1715644800, - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-large-32k-online) of this model has Internet access.", - "id": "perplexity/llama-3-sonar-large-32k-chat", - "name": "Perplexity: Llama3 Sonar 70B", + "context_length": 8192, + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3-8b-instruct", + "name": "Meta: Llama 3 8B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000001", + "completion": "0.000000055", "image": "0", - "prompt": "0.000001", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 28000, - "created": 1715644800, - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-small-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "id": "perplexity/llama-3-sonar-small-32k-online", - "name": "Perplexity: Llama3 Sonar 8B Online", + "context_length": 16384, + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are extended-context endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). They may have higher prices._", + "id": "meta-llama/llama-3-8b-instruct:extended", + "name": "Meta: Llama 3 8B Instruct (extended)", "per_request_limits": null, "pricing": { - "completion": "0.0000002", + "completion": "0.000001125", "image": "0", - "prompt": "0.0000002", - "request": "0.005" + "prompt": "0.0000001875", + "request": "0" }, "top_provider": { - "context_length": 28000, + "context_length": 16384, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1715644800, - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-small-32k-online) of this model has Internet access.", - "id": "perplexity/llama-3-sonar-small-32k-chat", - "name": "Perplexity: Llama3 Sonar 8B", + "context_length": 8192, + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3-8b-instruct:free", + "name": "Meta: Llama 3 8B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000002", + "completion": "0", "image": "0", - "prompt": "0.0000002", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "none", + "instruct_type": "llama3", "modality": "text->text", "tokenizer": "Llama3" }, "context_length": 8192, - "created": 1715558400, - "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-guard-2-8b", - "name": "Meta: LlamaGuard 2 8B", + "created": 1713398400, + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are higher-throughput endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). They may have higher prices._", + "id": "meta-llama/llama-3-8b-instruct:nitro", + "name": "Meta: Llama 3 8B Instruct (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.00000018", + "completion": "0.000000162", "image": "0", - "prompt": "0.00000018", + "prompt": "0.000000162", "request": "0" }, "top_provider": { @@ -1730,122 +1778,122 @@ }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "instruct_type": "none", + "modality": "text->text", + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1715558400, - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", - "id": "openai/gpt-4o-2024-05-13", - "name": "OpenAI: GPT-4o (2024-05-13)", + "context_length": 131072, + "created": 1722556800, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.1-405b", + "name": "Meta: Llama 3.1 405B (base)", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.007225", - "prompt": "0.000005", + "completion": "0.000002", + "image": "0", + "prompt": "0.000002", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "instruct_type": "llama3", + "modality": "text->text", + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1715558400, - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", - "id": "openai/gpt-4o", - "name": "OpenAI: GPT-4o", + "context_length": 131072, + "created": 1721692800, + "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.1-405b-instruct", + "name": "Meta: Llama 3.1 405B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.007225", - "prompt": "0.000005", + "completion": "0.00000179", + "image": "0", + "prompt": "0.00000179", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32000, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1715558400, - "description": "GPT-4o Extended is an experimental variant of GPT-4o with an extended max output tokens. This model supports only text input to text output.\n\n_These are extended-context endpoints for [GPT-4o](/models/openai/gpt-4o). They may have higher prices._", - "id": "openai/gpt-4o:extended", - "name": "OpenAI: GPT-4o (extended)", + "context_length": 131072, + "created": 1721692800, + "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.1 405B Instruct](/models/meta-llama/llama-3.1-405b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.1-405b-instruct:free", + "name": "Meta: Llama 3.1 405B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.000018", - "image": "0.007225", - "prompt": "0.000006", + "completion": "0", + "image": "0", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 64000 + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Qwen" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1715212800, - "description": "Qwen1.5 72B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "id": "qwen/qwen-72b-chat", - "name": "Qwen 1.5 72B Chat", + "context_length": 131072, + "created": 1721692800, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 70B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.1-70b-instruct", + "name": "Meta: Llama 3.1 70B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00000081", + "completion": "0.0000003", "image": "0", - "prompt": "0.00000081", + "prompt": "0.0000003", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 131072, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "llama3", "modality": "text->text", - "tokenizer": "Qwen" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1715212800, - "description": "Qwen1.5 110B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "id": "qwen/qwen-110b-chat", - "name": "Qwen 1.5 110B Chat", + "context_length": 131072, + "created": 1721692800, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 70B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.1 70B Instruct](/models/meta-llama/llama-3.1-70b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.1-70b-instruct:free", + "name": "Meta: Llama 3.1 70B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.00000162", + "completion": "0", "image": "0", - "prompt": "0.00000162", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { @@ -1854,20 +1902,20 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 24576, - "created": 1714780800, - "description": "The NeverSleep team is back, with a Llama 3 8B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "neversleep/llama-3-lumimaid-8b", - "name": "Llama 3 Lumimaid 8B", + "context_length": 131072, + "created": 1721692800, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.1-8b-instruct", + "name": "Meta: Llama 3.1 8B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000001125", + "completion": "0.000000055", "image": "0", - "prompt": "0.0000001875", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 100000, "is_moderated": false, "max_completion_tokens": null } @@ -1878,70 +1926,70 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 24576, - "created": 1714780800, - "description": "The NeverSleep team is back, with a Llama 3 8B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are extended-context endpoints for [Llama 3 Lumimaid 8B](/models/neversleep/llama-3-lumimaid-8b). They may have higher prices._", - "id": "neversleep/llama-3-lumimaid-8b:extended", - "name": "Llama 3 Lumimaid 8B (extended)", + "context_length": 131072, + "created": 1721692800, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.1 8B Instruct](/models/meta-llama/llama-3.1-8b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.1-8b-instruct:free", + "name": "Meta: Llama 3.1 8B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.000001125", + "completion": "0", "image": "0", - "prompt": "0.0000001875", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 24576, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "alpaca", - "modality": "text->text", - "tokenizer": "Llama2" + "instruct_type": "llama3", + "modality": "text+image->text", + "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713657600, - "description": "Creative writing model, routed with permission. It's fast, it keeps the conversation going, and it stays in character.\n\nIf you submit a raw prompt, you can use Alpaca or Vicuna formats.", - "id": "sao10k/fimbulvetr-11b-v2", - "name": "Fimbulvetr 11B v2", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.2-11b-vision-instruct", + "name": "Meta: Llama 3.2 11B Vision Instruct", "per_request_limits": null, "pricing": { - "completion": "0.0000015", - "image": "0", - "prompt": "0.000000375", + "completion": "0.000000055", + "image": "0.000079475", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 131072, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { "instruct_type": "llama3", - "modality": "text->text", + "modality": "text+image->text", "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3-70b-instruct", - "name": "Meta: Llama 3 70B Instruct", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.2 11B Vision Instruct](/models/meta-llama/llama-3.2-11b-vision-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.2-11b-vision-instruct:free", + "name": "Meta: Llama 3.2 11B Vision Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0", "image": "0", - "prompt": "0.00000035", + "prompt": "0", "request": "0" }, "top_provider": { "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { @@ -1950,20 +1998,20 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are higher-throughput endpoints for [Llama 3 70B Instruct](/models/meta-llama/llama-3-70b-instruct). They may have higher prices._", - "id": "meta-llama/llama-3-70b-instruct:nitro", - "name": "Meta: Llama 3 70B Instruct (nitro)", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.2-1b-instruct", + "name": "Meta: Llama 3.2 1B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000000792", + "completion": "0.00000002", "image": "0", - "prompt": "0.000000792", + "prompt": "0.00000001", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 131072, "is_moderated": false, "max_completion_tokens": null } @@ -1974,11 +2022,11 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "meta-llama/llama-3-8b-instruct:free", - "name": "Meta: Llama 3 8B Instruct (free)", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.2 1B Instruct](/models/meta-llama/llama-3.2-1b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.2-1b-instruct:free", + "name": "Meta: Llama 3.2 1B Instruct (free)", "per_request_limits": null, "pricing": { "completion": "0", @@ -1998,20 +2046,20 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "id": "meta-llama/llama-3-8b-instruct", - "name": "Meta: Llama 3 8B Instruct", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.2-3b-instruct", + "name": "Meta: Llama 3.2 3B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.00000005", "image": "0", - "prompt": "0.000000055", + "prompt": "0.00000003", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 131072, "is_moderated": false, "max_completion_tokens": null } @@ -2022,116 +2070,116 @@ "modality": "text->text", "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are higher-throughput endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). They may have higher prices._", - "id": "meta-llama/llama-3-8b-instruct:nitro", - "name": "Meta: Llama 3 8B Instruct (nitro)", + "context_length": 131072, + "created": 1727222400, + "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).\n\n_These are free, rate-limited endpoints for [Llama 3.2 3B Instruct](/models/meta-llama/llama-3.2-3b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "meta-llama/llama-3.2-3b-instruct:free", + "name": "Meta: Llama 3.2 3B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.000000162", + "completion": "0", "image": "0", - "prompt": "0.000000162", + "prompt": "0", "request": "0" }, "top_provider": { "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": "llama3", - "modality": "text->text", + "modality": "text+image->text", "tokenizer": "Llama3" }, - "context_length": 16384, - "created": 1713398400, - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).\n\n_These are extended-context endpoints for [Llama 3 8B Instruct](/models/meta-llama/llama-3-8b-instruct). They may have higher prices._", - "id": "meta-llama/llama-3-8b-instruct:extended", - "name": "Meta: Llama 3 8B Instruct (extended)", + "context_length": 131072, + "created": 1727222400, + "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-3.2-90b-vision-instruct", + "name": "Meta: Llama 3.2 90B Vision Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000001125", - "image": "0", - "prompt": "0.0000001875", + "completion": "0.0000004", + "image": "0.00050575", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 16384, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "llama2", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama2" }, - "context_length": 65536, - "created": 1713312000, - "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", - "id": "mistralai/mixtral-8x22b-instruct", - "name": "Mistral: Mixtral 8x22B Instruct", + "context_length": 4096, + "created": 1687219200, + "description": "A 13 billion parameter language model from Meta, fine tuned for chat completions", + "id": "meta-llama/llama-2-13b-chat", + "name": "Meta: Llama v2 13B Chat", "per_request_limits": null, "pricing": { - "completion": "0.00000065", + "completion": "0.000000198", "image": "0", - "prompt": "0.00000065", + "prompt": "0.000000198", "request": "0" }, "top_provider": { - "context_length": 65536, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "vicuna", + "instruct_type": "none", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32000, - "created": 1713225600, - "description": "WizardLM-2 7B is the smaller variant of Microsoft AI's latest Wizard model. It is the fastest and achieves comparable performance with existing 10x larger opensource leading models\n\nIt is a finetune of [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct), using the same technique as [WizardLM-2 8x22B](/models/microsoft/wizardlm-2-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", - "id": "microsoft/wizardlm-2-7b", - "name": "WizardLM-2 7B", + "context_length": 8192, + "created": 1715558400, + "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "id": "meta-llama/llama-guard-2-8b", + "name": "Meta: LlamaGuard 2 8B", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.00000018", "image": "0", - "prompt": "0.000000055", + "prompt": "0.00000018", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "vicuna", + "instruct_type": "airoboros", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama2" }, - "context_length": 65536, - "created": 1713225600, - "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", - "id": "microsoft/wizardlm-2-8x22b", - "name": "WizardLM-2 8x22B", + "context_length": 4096, + "created": 1711065600, + "description": "A merge with a complex family tree, this model was crafted for roleplaying and storytelling. Midnight Rose is a successor to Rogue Rose and Aurora Nights and improves upon them both. It wants to produce lengthy output by default and is the best creative writing merge produced so far by sophosympatheia.\n\nDescending from earlier versions of Midnight Rose and [Wizard Tulu Dolphin 70B](https://huggingface.co/sophosympatheia/Wizard-Tulu-Dolphin-70B-v1.0), it inherits the best qualities of each.", + "id": "sophosympatheia/midnight-rose-70b", + "name": "Midnight Rose 70B", "per_request_limits": null, "pricing": { - "completion": "0.0000005", + "completion": "0.0000008", "image": "0", - "prompt": "0.0000005", + "prompt": "0.0000008", "request": "0" }, "top_provider": { - "context_length": 65536, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } @@ -2139,119 +2187,119 @@ { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Gemini" + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 4000000, - "created": 1712620800, - "description": "Google's latest multimodal model, supporting image and video in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", - "id": "google/gemini-pro-1.5", - "name": "Google: Gemini Pro 1.5", + "context_length": 128000, + "created": 1708905600, + "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt is fluent in English, French, Spanish, German, and Italian, with high grammatical accuracy, and its long context window allows precise information recall from large documents.", + "id": "mistralai/mistral-large", + "name": "Mistral Large", "per_request_limits": null, "pricing": { - "completion": "0.0000075", - "image": "0.00263", - "prompt": "0.0000025", + "completion": "0.000006", + "image": "0", + "prompt": "0.000002", "request": "0" }, "top_provider": { - "context_length": 4000000, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": null } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "GPT" + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 128000, - "created": 1712620800, - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", - "id": "openai/gpt-4-turbo", - "name": "OpenAI: GPT-4 Turbo", + "context_length": 32000, + "created": 1704844800, + "description": "This is Mistral AI's closed-source, medium-sided model. It's powered by a closed-source prototype and excels at reasoning, code, JSON, chat, and more. In benchmarks, it compares with many of the flagship models of other companies.", + "id": "mistralai/mistral-medium", + "name": "Mistral Medium", "per_request_limits": null, "pricing": { - "completion": "0.00003", - "image": "0.01445", - "prompt": "0.00001", + "completion": "0.0000081", + "image": "0", + "prompt": "0.00000275", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32000, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Cohere" + "tokenizer": "Mistral" }, - "context_length": 128000, - "created": 1712188800, - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r-plus", - "name": "Cohere: Command R+", + "context_length": 32000, + "created": 1722556800, + "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", + "id": "nothingiisreal/mn-celeste-12b", + "name": "Mistral Nemo 12B Celeste", "per_request_limits": null, "pricing": { - "completion": "0.000015", + "completion": "0.0000015", "image": "0", - "prompt": "0.000003", + "prompt": "0.0000015", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 32000, "is_moderated": false, - "max_completion_tokens": 4000 + "max_completion_tokens": null } }, { "architecture": { "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Other" + "tokenizer": "Mistral" }, - "context_length": 32768, - "created": 1711670400, - "description": "DBRX is a new open source large language model developed by Databricks. At 132B, it outperforms existing open source LLMs like Llama 2 70B and [Mixtral-8x7b](/models/mistralai/mixtral-8x7b) on standard industry benchmarks for language understanding, programming, math, and logic.\n\nIt uses a fine-grained mixture-of-experts (MoE) architecture. 36B parameters are active on any input. It was pre-trained on 12T tokens of text and code data. Compared to other open MoE models like Mixtral-8x7B and Grok-1, DBRX is fine-grained, meaning it uses a larger number of smaller experts.\n\nSee the launch announcement and benchmark results [here](https://www.databricks.com/blog/introducing-dbrx-new-state-art-open-llm).\n\n#moe", - "id": "databricks/dbrx-instruct", - "name": "Databricks: DBRX 132B Instruct", + "context_length": 12000, + "created": 1723507200, + "description": "Starcannon 12B is a creative roleplay and story writing model, using [nothingiisreal/mn-celeste-12b](https://openrouter.ai/models/nothingiisreal/mn-celeste-12b) as a base and [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", + "id": "aetherwiing/mn-starcannon-12b", + "name": "Mistral Nemo 12B Starcannon", "per_request_limits": null, "pricing": { - "completion": "0.00000108", + "completion": "0.000002", "image": "0", - "prompt": "0.00000108", + "prompt": "0.000002", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 12000, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "airoboros", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Mistral" }, - "context_length": 4096, - "created": 1711065600, - "description": "A merge with a complex family tree, this model was crafted for roleplaying and storytelling. Midnight Rose is a successor to Rogue Rose and Aurora Nights and improves upon them both. It wants to produce lengthy output by default and is the best creative writing merge produced so far by sophosympatheia.\n\nDescending from earlier versions of Midnight Rose and [Wizard Tulu Dolphin 70B](https://huggingface.co/sophosympatheia/Wizard-Tulu-Dolphin-70B-v1.0), it inherits the best qualities of each.", - "id": "sophosympatheia/midnight-rose-70b", - "name": "Midnight Rose 70B", + "context_length": 32000, + "created": 1704844800, + "description": "Cost-efficient, fast, and reliable option for use cases such as translation, summarization, and sentiment analysis.", + "id": "mistralai/mistral-small", + "name": "Mistral Small", "per_request_limits": null, "pricing": { - "completion": "0.0000008", + "completion": "0.0000006", "image": "0", - "prompt": "0.0000008", + "prompt": "0.0000002", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 32000, "is_moderated": false, "max_completion_tokens": null } @@ -2260,210 +2308,210 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Cohere" + "tokenizer": "Mistral" }, - "context_length": 128000, - "created": 1710374400, - "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command-r", - "name": "Cohere: Command R", + "context_length": 32000, + "created": 1704844800, + "description": "This model is currently powered by Mistral-7B-v0.2, and incorporates a \"better\" fine-tuning than [Mistral 7B](/models/mistralai/mistral-7b-instruct-v0.1), inspired by community work. It's best used for large batch processing tasks where cost is a significant factor but reasoning capabilities are not crucial.", + "id": "mistralai/mistral-tiny", + "name": "Mistral Tiny", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0.00000025", "image": "0", - "prompt": "0.0000005", + "prompt": "0.00000025", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 32000, "is_moderated": false, - "max_completion_tokens": 4000 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "mistral", "modality": "text->text", - "tokenizer": "Cohere" + "tokenizer": "Mistral" }, - "context_length": 4096, - "created": 1710374400, - "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).", - "id": "cohere/command", - "name": "Cohere: Command", + "context_length": 256000, + "created": 1721347200, + "description": "A 7.3B parameter Mamba-based model designed for code and reasoning tasks.\n\n- Linear time inference, allowing for theoretically infinite sequence lengths\n- 256k token context window\n- Optimized for quick responses, especially beneficial for code productivity\n- Performs comparably to state-of-the-art transformer models in code and reasoning tasks\n- Available under the Apache 2.0 license for free use, modification, and distribution", + "id": "mistralai/codestral-mamba", + "name": "Mistral: Codestral Mamba", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.00000025", "image": "0", - "prompt": "0.000001", + "prompt": "0.00000025", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 256000, "is_moderated": false, - "max_completion_tokens": 4000 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1710288000, - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", - "id": "anthropic/claude-3-haiku", - "name": "Anthropic: Claude 3 Haiku", + "context_length": 32768, + "created": 1716768000, + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", + "id": "mistralai/mistral-7b-instruct", + "name": "Mistral: Mistral 7B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00000125", - "image": "0.0004", - "prompt": "0.00000025", + "completion": "0.000000055", + "image": "0", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1710288000, - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-haiku) variant._", - "id": "anthropic/claude-3-haiku:beta", - "name": "Anthropic: Claude 3 Haiku (self-moderated)", + "context_length": 32768, + "created": 1716768000, + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*\n\n_These are free, rate-limited endpoints for [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "mistralai/mistral-7b-instruct:free", + "name": "Mistral: Mistral 7B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.00000125", - "image": "0.0004", - "prompt": "0.00000025", + "completion": "0", + "image": "0", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 8192, "is_moderated": false, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1709596800, - "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "id": "anthropic/claude-3-sonnet", - "name": "Anthropic: Claude 3 Sonnet", + "context_length": 32768, + "created": 1716768000, + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*\n\n_These are higher-throughput endpoints for [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct). They may have higher prices._", + "id": "mistralai/mistral-7b-instruct:nitro", + "name": "Mistral: Mistral 7B Instruct (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.000015", - "image": "0.0048", - "prompt": "0.000003", + "completion": "0.00000007", + "image": "0", + "prompt": "0.00000007", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1709596800, - "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-sonnet) variant._", - "id": "anthropic/claude-3-sonnet:beta", - "name": "Anthropic: Claude 3 Sonnet (self-moderated)", - "per_request_limits": null, - "pricing": { - "completion": "0.000015", - "image": "0.0048", - "prompt": "0.000003", + "context_length": 4096, + "created": 1695859200, + "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", + "id": "mistralai/mistral-7b-instruct-v0.1", + "name": "Mistral: Mistral 7B Instruct v0.1", + "per_request_limits": null, + "pricing": { + "completion": "0.00000018", + "image": "0", + "prompt": "0.00000018", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1709596800, - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "id": "anthropic/claude-3-opus", - "name": "Anthropic: Claude 3 Opus", + "context_length": 32768, + "created": 1703721600, + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", + "id": "mistralai/mistral-7b-instruct-v0.2", + "name": "Mistral: Mistral 7B Instruct v0.2", "per_request_limits": null, "pricing": { - "completion": "0.000075", - "image": "0.024", - "prompt": "0.000015", + "completion": "0.00000018", + "image": "0", + "prompt": "0.00000018", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Claude" + "instruct_type": "mistral", + "modality": "text->text", + "tokenizer": "Mistral" }, - "context_length": 200000, - "created": 1709596800, - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-3-opus) variant._", - "id": "anthropic/claude-3-opus:beta", - "name": "Anthropic: Claude 3 Opus (self-moderated)", + "context_length": 32768, + "created": 1716768000, + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct v0.2](/models/mistralai/mistral-7b-instruct-v0.2), with the following changes:\n\n- Extended vocabulary to 32768\n- Supports v3 Tokenizer\n- Supports function calling\n\nNOTE: Support for function calling depends on the provider.", + "id": "mistralai/mistral-7b-instruct-v0.3", + "name": "Mistral: Mistral 7B Instruct v0.3", "per_request_limits": null, "pricing": { - "completion": "0.000075", - "image": "0.024", - "prompt": "0.000015", + "completion": "0.000000055", + "image": "0", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "mistral", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 128000, - "created": 1708905600, - "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt is fluent in English, French, Spanish, German, and Italian, with high grammatical accuracy, and its long context window allows precise information recall from large documents.", - "id": "mistralai/mistral-large", - "name": "Mistral Large", + "created": 1721347200, + "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", + "id": "mistralai/mistral-nemo", + "name": "Mistral: Mistral Nemo", "per_request_limits": null, "pricing": { - "completion": "0.000009", + "completion": "0.00000013", "image": "0", - "prompt": "0.000003", + "prompt": "0.00000013", "request": "0" }, "top_provider": { @@ -2474,68 +2522,68 @@ }, { "architecture": { - "instruct_type": null, + "instruct_type": "mistral", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Mistral" }, - "context_length": 128000, - "created": 1706140800, - "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", - "id": "openai/gpt-4-turbo-preview", - "name": "OpenAI: GPT-4 Turbo Preview", + "context_length": 65536, + "created": 1713312000, + "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", + "id": "mistralai/mixtral-8x22b-instruct", + "name": "Mistral: Mixtral 8x22B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00003", + "completion": "0.0000009", "image": "0", - "prompt": "0.00001", + "prompt": "0.0000009", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 65536, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "GPT" + "instruct_type": "mistral", + "modality": "text+image->text", + "tokenizer": "Mistral" }, - "context_length": 4095, - "created": 1706140800, - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", - "id": "openai/gpt-3.5-turbo-0613", - "name": "OpenAI: GPT-3.5 Turbo (older v0613)", + "context_length": 4096, + "created": 1725926400, + "description": "The first image to text model from Mistral AI. Its weight was launched via torrent per their tradition: https://x.com/mistralai/status/1833758285167722836", + "id": "mistralai/pixtral-12b", + "name": "Mistral: Pixtral 12B", "per_request_limits": null, "pricing": { - "completion": "0.000002", - "image": "0", - "prompt": "0.000001", + "completion": "0.0000001", + "image": "0.0001445", + "prompt": "0.0000001", "request": "0" }, "top_provider": { - "context_length": 4095, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 4096, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "none", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 32768, - "created": 1705363200, - "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", - "id": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "name": "Nous: Hermes 2 Mixtral 8x7B DPO", + "created": 1702166400, + "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI. Incorporates 8 experts (feed-forward networks) for a total of 47B parameters. Base model (not fine-tuned for instructions) - see [Mixtral 8x7B Instruct](/models/mistralai/mixtral-8x7b-instruct) for an instruct-tuned model.\n\n#moe", + "id": "mistralai/mixtral-8x7b", + "name": "Mixtral 8x7B (base)", "per_request_limits": null, "pricing": { - "completion": "0.00000045", + "completion": "0.00000054", "image": "0", - "prompt": "0.00000045", + "prompt": "0.00000054", "request": "0" }, "top_provider": { @@ -2546,72 +2594,72 @@ }, { "architecture": { - "instruct_type": null, + "instruct_type": "mistral", "modality": "text->text", "tokenizer": "Mistral" }, - "context_length": 32000, - "created": 1704844800, - "description": "This is Mistral AI's closed-source, medium-sided model. It's powered by a closed-source prototype and excels at reasoning, code, JSON, chat, and more. In benchmarks, it compares with many of the flagship models of other companies.", - "id": "mistralai/mistral-medium", - "name": "Mistral Medium", + "context_length": 32768, + "created": 1702166400, + "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe", + "id": "mistralai/mixtral-8x7b-instruct", + "name": "Mixtral 8x7B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.0000081", + "completion": "0.00000024", "image": "0", - "prompt": "0.0000027", + "prompt": "0.00000024", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "mistral", "modality": "text->text", "tokenizer": "Mistral" }, - "context_length": 32000, - "created": 1704844800, - "description": "This model is currently powered by Mixtral-8X7B-v0.1, a sparse mixture of experts model with 12B active parameters. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.\n#moe", - "id": "mistralai/mistral-small", - "name": "Mistral Small", + "context_length": 32768, + "created": 1702166400, + "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe\n\n_These are higher-throughput endpoints for [Mixtral 8x7B Instruct](/models/mistralai/mixtral-8x7b-instruct). They may have higher prices._", + "id": "mistralai/mixtral-8x7b-instruct:nitro", + "name": "Mixtral 8x7B Instruct (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.000006", + "completion": "0.00000054", "image": "0", - "prompt": "0.000002", + "prompt": "0.00000054", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama2" }, - "context_length": 32000, - "created": 1704844800, - "description": "This model is currently powered by Mistral-7B-v0.2, and incorporates a \"better\" fine-tuning than [Mistral 7B](/models/mistralai/mistral-7b-instruct-v0.1), inspired by community work. It's best used for large batch processing tasks where cost is a significant factor but reasoning capabilities are not crucial.", - "id": "mistralai/mistral-tiny", - "name": "Mistral Tiny", + "context_length": 4096, + "created": 1688256000, + "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", + "id": "gryphe/mythomax-l2-13b", + "name": "MythoMax 13B", "per_request_limits": null, "pricing": { - "completion": "0.00000025", + "completion": "0.0000001", "image": "0", - "prompt": "0.00000025", + "prompt": "0.0000001", "request": "0" }, "top_provider": { - "context_length": 32000, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } @@ -2622,40 +2670,40 @@ "modality": "text->text", "tokenizer": "Llama2" }, - "context_length": 4096, - "created": 1704412800, - "description": "A 75/25 merge of [Chronos 13b v2](https://huggingface.co/elinas/chronos-13b-v2) and [Nous Hermes Llama2 13b](/models/nousresearch/nous-hermes-llama2-13b). This offers the imaginative writing style of Chronos while retaining coherency. Outputs are long and use exceptional prose. #merge", - "id": "austism/chronos-hermes-13b", - "name": "Chronos Hermes 13B v2", + "context_length": 8192, + "created": 1688256000, + "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge\n\n_These are extended-context endpoints for [MythoMax 13B](/models/gryphe/mythomax-l2-13b). They may have higher prices._", + "id": "gryphe/mythomax-l2-13b:extended", + "name": "MythoMax 13B (extended)", "per_request_limits": null, "pricing": { - "completion": "0.00000013", + "completion": "0.000001125", "image": "0", - "prompt": "0.00000013", + "prompt": "0.000001125", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 400 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Yi" + "tokenizer": "Llama2" }, "context_length": 4096, - "created": 1704153600, - "description": "Nous Hermes 2 Yi 34B was trained on 1,000,000 entries of primarily GPT-4 generated data, as well as other high quality data from open datasets across the AI landscape.\n\nNous-Hermes 2 on Yi 34B outperforms all Nous-Hermes & Open-Hermes models of the past, achieving new heights in all benchmarks for a Nous Research LLM as well as surpassing many popular finetunes.", - "id": "nousresearch/nous-hermes-yi-34b", - "name": "Nous: Hermes 2 Yi 34B", + "created": 1688256000, + "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge\n\n_These are higher-throughput endpoints for [MythoMax 13B](/models/gryphe/mythomax-l2-13b). They may have higher prices._", + "id": "gryphe/mythomax-l2-13b:nitro", + "name": "MythoMax 13B (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.00000072", + "completion": "0.0000002", "image": "0", - "prompt": "0.00000072", + "prompt": "0.0000002", "request": "0" }, "top_provider": { @@ -2666,135 +2714,111 @@ }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "alpaca", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 32768, - "created": 1703721600, - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", - "id": "mistralai/mistral-7b-instruct-v0.2", - "name": "Mistral: Mistral 7B Instruct v0.2", + "created": 1701907200, + "description": "From the creator of [MythoMax](/models/gryphe/mythomax-l2-13b), merges a suite of models to reduce word anticipation, ministrations, and other undesirable words in ChatGPT roleplaying data.\n\nIt combines [Neural Chat 7B](/models/intel/neural-chat-7b), Airoboros 7b, [Toppy M 7B](/models/undi95/toppy-m-7b), [Zepher 7b beta](/models/huggingfaceh4/zephyr-7b-beta), [Nous Capybara 34B](/models/nousresearch/nous-capybara-34b), [OpenHeremes 2.5](/models/teknium/openhermes-2.5-mistral-7b), and many others.\n\n#merge", + "id": "gryphe/mythomist-7b", + "name": "MythoMist 7B", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.000000375", "image": "0", - "prompt": "0.000000055", + "prompt": "0.000000375", "request": "0" }, "top_provider": { "context_length": 32768, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": "alpaca", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 32768, - "created": 1703116800, - "description": "This is a 16k context fine-tune of [Mixtral-8x7b](/models/mistralai/mixtral-8x7b). It excels in coding tasks due to extensive training with coding data and is known for its obedience, although it lacks DPO tuning.\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored", - "id": "cognitivecomputations/dolphin-mixtral-8x7b", - "name": "Dolphin 2.6 Mixtral 8x7B 🐬", - "per_request_limits": null, - "pricing": { - "completion": "0.0000005", - "image": "0", - "prompt": "0.0000005", - "request": "0" - }, - "top_provider": { - "context_length": 32768, - "is_moderated": false, - "max_completion_tokens": null - } - }, - { - "architecture": { - "instruct_type": null, - "modality": "text+image->text", - "tokenizer": "Gemini" - }, - "context_length": 65536, - "created": 1702425600, - "description": "Google's flagship multimodal model, supporting image and video in text or chat prompts for a text or code response.\n\nSee the benchmarks and prompting guidelines from [Deepmind](https://deepmind.google/technologies/gemini/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal", - "id": "google/gemini-pro-vision", - "name": "Google: Gemini Pro Vision 1.0", + "created": 1701907200, + "description": "From the creator of [MythoMax](/models/gryphe/mythomax-l2-13b), merges a suite of models to reduce word anticipation, ministrations, and other undesirable words in ChatGPT roleplaying data.\n\nIt combines [Neural Chat 7B](/models/intel/neural-chat-7b), Airoboros 7b, [Toppy M 7B](/models/undi95/toppy-m-7b), [Zepher 7b beta](/models/huggingfaceh4/zephyr-7b-beta), [Nous Capybara 34B](/models/nousresearch/nous-capybara-34b), [OpenHeremes 2.5](/models/teknium/openhermes-2.5-mistral-7b), and many others.\n\n#merge\n\n_These are free, rate-limited endpoints for [MythoMist 7B](/models/gryphe/mythomist-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "gryphe/mythomist-7b:free", + "name": "MythoMist 7B (free)", "per_request_limits": null, "pricing": { - "completion": "0.000000375", - "image": "0.0025", - "prompt": "0.000000125", + "completion": "0", + "image": "0", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 65536, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 8192 + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Gemini" + "tokenizer": "Llama2" }, - "context_length": 131040, - "created": 1702425600, - "description": "Google's flagship text generation model. Designed to handle natural language tasks, multiturn text and code chat, and code generation.\n\nSee the benchmarks and prompting guidelines from [Deepmind](https://deepmind.google/technologies/gemini/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", - "id": "google/gemini-pro", - "name": "Google: Gemini Pro 1.0", + "context_length": 8192, + "created": 1700956800, + "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", + "id": "neversleep/noromaid-20b", + "name": "Noromaid 20B", "per_request_limits": null, "pricing": { - "completion": "0.000000375", - "image": "0.0025", - "prompt": "0.000000125", + "completion": "0.00000225", + "image": "0", + "prompt": "0.0000015", "request": "0" }, "top_provider": { - "context_length": 131040, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama2" }, - "context_length": 32768, - "created": 1702166400, - "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe", - "id": "mistralai/mixtral-8x7b-instruct", - "name": "Mixtral 8x7B Instruct", + "context_length": 4096, + "created": 1692489600, + "description": "A state-of-the-art language model fine-tuned on over 300k instructions by Nous Research, with Teknium and Emozilla leading the fine tuning process.", + "id": "nousresearch/nous-hermes-llama2-13b", + "name": "Nous: Hermes 13B", "per_request_limits": null, "pricing": { - "completion": "0.00000024", + "completion": "0.00000017", "image": "0", - "prompt": "0.00000024", + "prompt": "0.00000017", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": "chatml", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 32768, - "created": 1702166400, - "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe\n\n_These are higher-throughput endpoints for [Mixtral 8x7B Instruct](/models/mistralai/mixtral-8x7b-instruct). They may have higher prices._", - "id": "mistralai/mixtral-8x7b-instruct:nitro", - "name": "Mixtral 8x7B Instruct (nitro)", + "created": 1705363200, + "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", + "id": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "name": "Nous: Hermes 2 Mixtral 8x7B DPO", "per_request_limits": null, "pricing": { "completion": "0.00000054", @@ -2810,111 +2834,111 @@ }, { "architecture": { - "instruct_type": "none", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1702166400, - "description": "A pretrained generative Sparse Mixture of Experts, by Mistral AI. Incorporates 8 experts (feed-forward networks) for a total of 47B parameters. Base model (not fine-tuned for instructions) - see [Mixtral 8x7B Instruct](/models/mistralai/mixtral-8x7b-instruct) for an instruct-tuned model.\n\n#moe", - "id": "mistralai/mixtral-8x7b", - "name": "Mixtral 8x7B (base)", + "context_length": 16384, + "created": 1720656000, + "description": "An experimental merge model based on Llama 3, exhibiting a very distinctive style of writing. It combines the the best of [Meta's Llama 3 8B](https://openrouter.ai/models/meta-llama/llama-3-8b-instruct) and Nous Research's [Hermes 2 Pro](https://openrouter.ai/models/nousresearch/hermes-2-pro-llama-3-8b).\n\nHermes-2 Θ (theta) was specifically designed with a few capabilities in mind: executing function calls, generating JSON output, and most remarkably, demonstrating metacognitive abilities (contemplating the nature of thought and recognizing the diversity of cognitive processes among individuals).", + "id": "nousresearch/hermes-2-theta-llama-3-8b", + "name": "Nous: Hermes 2 Theta 8B", "per_request_limits": null, "pricing": { - "completion": "0.00000054", + "completion": "0.000001125", "image": "0", - "prompt": "0.00000054", + "prompt": "0.0000001875", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 16384, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Yi" }, - "context_length": 32768, - "created": 1702080000, - "description": "This is the chat model variant of the [StripedHyena series](/models?q=stripedhyena) developed by Together in collaboration with Nous Research.\n\nStripedHyena uses a new architecture that competes with traditional Transformers, particularly in long-context data processing. It combines attention mechanisms with gated convolutions for improved speed, efficiency, and scaling. This model marks a significant advancement in AI architecture for sequence modeling tasks.", - "id": "togethercomputer/stripedhyena-nous-7b", - "name": "StripedHyena Nous 7B", + "context_length": 4096, + "created": 1704153600, + "description": "Nous Hermes 2 Yi 34B was trained on 1,000,000 entries of primarily GPT-4 generated data, as well as other high quality data from open datasets across the AI landscape.\n\nNous-Hermes 2 on Yi 34B outperforms all Nous-Hermes & Open-Hermes models of the past, achieving new heights in all benchmarks for a Nous Research LLM as well as surpassing many popular finetunes.", + "id": "nousresearch/nous-hermes-yi-34b", + "name": "Nous: Hermes 2 Yi 34B", "per_request_limits": null, "pricing": { - "completion": "0.00000018", + "completion": "0.00000072", "image": "0", - "prompt": "0.00000018", + "prompt": "0.00000072", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 4096, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1701907200, - "description": "From the creator of [MythoMax](/models/gryphe/mythomax-l2-13b), merges a suite of models to reduce word anticipation, ministrations, and other undesirable words in ChatGPT roleplaying data.\n\nIt combines [Neural Chat 7B](/models/intel/neural-chat-7b), Airoboros 7b, [Toppy M 7B](/models/undi95/toppy-m-7b), [Zepher 7b beta](/models/huggingfaceh4/zephyr-7b-beta), [Nous Capybara 34B](/models/nousresearch/nous-capybara-34b), [OpenHeremes 2.5](/models/teknium/openhermes-2.5-mistral-7b), and many others.\n\n#merge\n\n_These are free, rate-limited endpoints for [MythoMist 7B](/models/gryphe/mythomist-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "gryphe/mythomist-7b:free", - "name": "MythoMist 7B (free)", + "context_length": 131072, + "created": 1723766400, + "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.", + "id": "nousresearch/hermes-3-llama-3.1-405b", + "name": "Nous: Hermes 3 405B Instruct", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.0000045", "image": "0", - "prompt": "0", + "prompt": "0.0000045", "request": "0" }, "top_provider": { "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 32768, - "created": 1701907200, - "description": "From the creator of [MythoMax](/models/gryphe/mythomax-l2-13b), merges a suite of models to reduce word anticipation, ministrations, and other undesirable words in ChatGPT roleplaying data.\n\nIt combines [Neural Chat 7B](/models/intel/neural-chat-7b), Airoboros 7b, [Toppy M 7B](/models/undi95/toppy-m-7b), [Zepher 7b beta](/models/huggingfaceh4/zephyr-7b-beta), [Nous Capybara 34B](/models/nousresearch/nous-capybara-34b), [OpenHeremes 2.5](/models/teknium/openhermes-2.5-mistral-7b), and many others.\n\n#merge", - "id": "gryphe/mythomist-7b", - "name": "MythoMist 7B", + "context_length": 128000, + "created": 1723766400, + "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.\n\n_These are extended-context endpoints for [Hermes 3 405B Instruct](/models/nousresearch/hermes-3-llama-3.1-405b). They may have higher prices._", + "id": "nousresearch/hermes-3-llama-3.1-405b:extended", + "name": "Nous: Hermes 3 405B Instruct (extended)", "per_request_limits": null, "pricing": { - "completion": "0.000000375", + "completion": "0.0000045", "image": "0", - "prompt": "0.000000375", + "prompt": "0.0000045", "request": "0" }, "top_provider": { - "context_length": 32768, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "openchat", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1701129600, - "description": "OpenChat 7B is a library of open-source language models, fine-tuned with \"C-RLFT (Conditioned Reinforcement Learning Fine-Tuning)\" - a strategy inspired by offline reinforcement learning. It has been trained on mixed-quality data without preference labels.\n\n- For OpenChat fine-tuned on Mistral 7B, check out [OpenChat 7B](/models/openchat/openchat-7b).\n- For OpenChat fine-tuned on Llama 8B, check out [OpenChat 8B](/models/openchat/openchat-8b).\n\n#open-source\n\n_These are free, rate-limited endpoints for [OpenChat 3.5 7B](/models/openchat/openchat-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "openchat/openchat-7b:free", - "name": "OpenChat 3.5 7B (free)", + "context_length": 131072, + "created": 1723766400, + "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.\n\n_These are free, rate-limited endpoints for [Hermes 3 405B Instruct](/models/nousresearch/hermes-3-llama-3.1-405b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "nousresearch/hermes-3-llama-3.1-405b:free", + "name": "Nous: Hermes 3 405B Instruct (free)", "per_request_limits": null, "pricing": { "completion": "0", @@ -2930,96 +2954,96 @@ }, { "architecture": { - "instruct_type": "openchat", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1701129600, - "description": "OpenChat 7B is a library of open-source language models, fine-tuned with \"C-RLFT (Conditioned Reinforcement Learning Fine-Tuning)\" - a strategy inspired by offline reinforcement learning. It has been trained on mixed-quality data without preference labels.\n\n- For OpenChat fine-tuned on Mistral 7B, check out [OpenChat 7B](/models/openchat/openchat-7b).\n- For OpenChat fine-tuned on Llama 8B, check out [OpenChat 8B](/models/openchat/openchat-8b).\n\n#open-source", - "id": "openchat/openchat-7b", - "name": "OpenChat 3.5 7B", + "context_length": 131072, + "created": 1723939200, + "description": "Hermes 3 is a generalist language model with many improvements over [Hermes 2](/models/nousresearch/nous-hermes-2-mistral-7b-dpo), including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 70B is a competitive, if not superior finetune of the [Llama-3.1 70B foundation model](/models/meta-llama/llama-3.1-70b-instruct), focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.", + "id": "nousresearch/hermes-3-llama-3.1-70b", + "name": "Nous: Hermes 3 70B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.0000004", "image": "0", - "prompt": "0.000000055", + "prompt": "0.0000004", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 12288, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Llama3" }, "context_length": 8192, - "created": 1700956800, - "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", - "id": "neversleep/noromaid-20b", - "name": "Noromaid 20B", + "created": 1716768000, + "description": "Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.", + "id": "nousresearch/hermes-2-pro-llama-3-8b", + "name": "NousResearch: Hermes 2 Pro - Llama-3 8B", "per_request_limits": null, "pricing": { - "completion": "0.00000225", + "completion": "0.00000014", "image": "0", - "prompt": "0.0000015", + "prompt": "0.00000014", "request": "0" }, "top_provider": { "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "claude", - "modality": "text->text", - "tokenizer": "Claude" + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" }, - "context_length": 100000, - "created": 1700611200, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", - "id": "anthropic/claude-instant-1.1", - "name": "Anthropic: Claude Instant v1.1", + "context_length": 128000, + "created": 1723593600, + "description": "Dynamic model continuously updated to the current version of [GPT-4o](/models/openai/gpt-4o) in ChatGPT. Intended for research and evaluation.\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "openai/chatgpt-4o-latest", + "name": "OpenAI: ChatGPT-4o", "per_request_limits": null, "pricing": { - "completion": "0.0000024", - "image": "0", - "prompt": "0.0000008", + "completion": "0.000015", + "image": "0.007225", + "prompt": "0.000005", "request": "0" }, "top_provider": { - "context_length": 100000, + "context_length": 128000, "is_moderated": true, - "max_completion_tokens": 2048 + "max_completion_tokens": 16384 } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "GPT" }, - "context_length": 200000, - "created": 1700611200, - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", - "id": "anthropic/claude-2.1", - "name": "Anthropic: Claude v2.1", + "context_length": 16385, + "created": 1685232000, + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "id": "openai/gpt-3.5-turbo", + "name": "OpenAI: GPT-3.5 Turbo", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.0000015", "image": "0", - "prompt": "0.000008", + "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 16385, "is_moderated": true, "max_completion_tokens": 4096 } @@ -3028,23 +3052,23 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "GPT" }, - "context_length": 200000, - "created": 1700611200, - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2.1) variant._", - "id": "anthropic/claude-2.1:beta", - "name": "Anthropic: Claude v2.1 (self-moderated)", + "context_length": 4095, + "created": 1685232000, + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "id": "openai/gpt-3.5-turbo-0301", + "name": "OpenAI: GPT-3.5 Turbo (older v0301)", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.000002", "image": "0", - "prompt": "0.000008", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": false, + "context_length": 4095, + "is_moderated": true, "max_completion_tokens": 4096 } }, @@ -3052,22 +3076,22 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "GPT" }, - "context_length": 200000, - "created": 1700611200, - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", - "id": "anthropic/claude-2", - "name": "Anthropic: Claude v2", + "context_length": 4095, + "created": 1706140800, + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "id": "openai/gpt-3.5-turbo-0613", + "name": "OpenAI: GPT-3.5 Turbo (older v0613)", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.000002", "image": "0", - "prompt": "0.000008", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 200000, + "context_length": 4095, "is_moderated": true, "max_completion_tokens": 4096 } @@ -3076,23 +3100,71 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "GPT" }, - "context_length": 200000, - "created": 1700611200, - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2) variant._", - "id": "anthropic/claude-2:beta", - "name": "Anthropic: Claude v2 (self-moderated)", + "context_length": 16385, + "created": 1693180800, + "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", + "id": "openai/gpt-3.5-turbo-16k", + "name": "OpenAI: GPT-3.5 Turbo 16k", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.000004", "image": "0", - "prompt": "0.000008", + "prompt": "0.000003", "request": "0" }, "top_provider": { - "context_length": 200000, - "is_moderated": false, + "context_length": 16385, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 16385, + "created": 1685232000, + "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", + "id": "openai/gpt-3.5-turbo-0125", + "name": "OpenAI: GPT-3.5 Turbo 16k", + "per_request_limits": null, + "pricing": { + "completion": "0.0000015", + "image": "0", + "prompt": "0.0000005", + "request": "0" + }, + "top_provider": { + "context_length": 16385, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 16385, + "created": 1699228800, + "description": "An older GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.", + "id": "openai/gpt-3.5-turbo-1106", + "name": "OpenAI: GPT-3.5 Turbo 16k (older v1106)", + "per_request_limits": null, + "pricing": { + "completion": "0.000002", + "image": "0", + "prompt": "0.000001", + "request": "0" + }, + "top_provider": { + "context_length": 16385, + "is_moderated": true, "max_completion_tokens": 4096 } }, @@ -3100,162 +3172,522 @@ "architecture": { "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "GPT" }, - "context_length": 4096, - "created": 1700438400, - "description": "A continuation of [OpenHermes 2 model](/models/teknium/openhermes-2-mistral-7b), trained on additional code datasets.\nPotentially the most interesting finding from training on a good ratio (est. of around 7-14% of the total dataset) of code instruction was that it has boosted several non-code benchmarks, including TruthfulQA, AGIEval, and GPT4All suite. It did however reduce BigBench benchmark score, but the net gain overall is significant.", - "id": "teknium/openhermes-2.5-mistral-7b", - "name": "OpenHermes 2.5 Mistral 7B", + "context_length": 4095, + "created": 1695859200, + "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", + "id": "openai/gpt-3.5-turbo-instruct", + "name": "OpenAI: GPT-3.5 Turbo Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00000017", + "completion": "0.000002", "image": "0", - "prompt": "0.00000017", + "prompt": "0.0000015", "request": "0" }, "top_provider": { - "context_length": 4096, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 4095, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { "instruct_type": null, - "modality": "text+image->text", + "modality": "text->text", "tokenizer": "GPT" }, - "context_length": 128000, - "created": 1699833600, - "description": "Ability to understand images, in addition to all other [GPT-4 Turbo capabilties](/models/openai/gpt-4-turbo). Training data: up to Apr 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.\n\n#multimodal", - "id": "openai/gpt-4-vision-preview", - "name": "OpenAI: GPT-4 Vision", + "context_length": 8191, + "created": 1685232000, + "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", + "id": "openai/gpt-4", + "name": "OpenAI: GPT-4", "per_request_limits": null, "pricing": { - "completion": "0.00003", - "image": "0.01445", - "prompt": "0.00001", + "completion": "0.00006", + "image": "0", + "prompt": "0.00003", "request": "0" }, "top_provider": { - "context_length": 128000, + "context_length": 8191, "is_moderated": true, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "airoboros", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "GPT" }, - "context_length": 4096, - "created": 1699747200, - "description": "A Mythomax/MLewd_13B-style merge of selected 70B models.\nA multi-model merge of several LLaMA2 70B finetunes for roleplaying and creative work. The goal was to create a model that combines creativity with intelligence for an enhanced experience.\n\n#merge #uncensored", - "id": "lizpreciatior/lzlv-70b-fp16-hf", - "name": "lzlv 70B", + "context_length": 8191, + "created": 1685232000, + "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", + "id": "openai/gpt-4-0314", + "name": "OpenAI: GPT-4 (older v0314)", "per_request_limits": null, "pricing": { - "completion": "0.0000004", + "completion": "0.00006", "image": "0", - "prompt": "0.00000035", + "prompt": "0.00003", "request": "0" }, "top_provider": { - "context_length": 4096, - "is_moderated": false, - "max_completion_tokens": null + "context_length": 8191, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "airoboros", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "GPT" }, - "context_length": 6144, - "created": 1699574400, - "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", - "id": "alpindale/goliath-120b", - "name": "Goliath 120B", + "context_length": 32767, + "created": 1693180800, + "description": "GPT-4-32k is an extended version of GPT-4, with the same capabilities but quadrupled context length, allowing for processing up to 40 pages of text in a single pass. This is particularly beneficial for handling longer content like interacting with PDFs without an external vector database. Training data: up to Sep 2021.", + "id": "openai/gpt-4-32k", + "name": "OpenAI: GPT-4 32k", "per_request_limits": null, "pricing": { - "completion": "0.000009375", + "completion": "0.00012", "image": "0", - "prompt": "0.000009375", + "prompt": "0.00006", "request": "0" }, "top_provider": { - "context_length": 6144, - "is_moderated": false, - "max_completion_tokens": 400 + "context_length": 32767, + "is_moderated": true, + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "GPT" }, - "context_length": 4096, - "created": 1699574400, - "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored\n\n_These are free, rate-limited endpoints for [Toppy M 7B](/models/undi95/toppy-m-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "undi95/toppy-m-7b:free", - "name": "Toppy M 7B (free)", + "context_length": 32767, + "created": 1693180800, + "description": "GPT-4-32k is an extended version of GPT-4, with the same capabilities but quadrupled context length, allowing for processing up to 40 pages of text in a single pass. This is particularly beneficial for handling longer content like interacting with PDFs without an external vector database. Training data: up to Sep 2021.", + "id": "openai/gpt-4-32k-0314", + "name": "OpenAI: GPT-4 32k (older v0314)", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.00012", + "image": "0", + "prompt": "0.00006", + "request": "0" + }, + "top_provider": { + "context_length": 32767, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1712620800, + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", + "id": "openai/gpt-4-turbo", + "name": "OpenAI: GPT-4 Turbo", + "per_request_limits": null, + "pricing": { + "completion": "0.00003", + "image": "0.01445", + "prompt": "0.00001", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1699228800, + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", + "id": "openai/gpt-4-1106-preview", + "name": "OpenAI: GPT-4 Turbo (older v1106)", + "per_request_limits": null, + "pricing": { + "completion": "0.00003", + "image": "0", + "prompt": "0.00001", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1706140800, + "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", + "id": "openai/gpt-4-turbo-preview", + "name": "OpenAI: GPT-4 Turbo Preview", + "per_request_limits": null, + "pricing": { + "completion": "0.00003", + "image": "0", + "prompt": "0.00001", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1699833600, + "description": "Ability to understand images, in addition to all other [GPT-4 Turbo capabilties](/models/openai/gpt-4-turbo). Training data: up to Apr 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.\n\n#multimodal", + "id": "openai/gpt-4-vision-preview", + "name": "OpenAI: GPT-4 Vision", + "per_request_limits": null, + "pricing": { + "completion": "0.00003", + "image": "0.01445", + "prompt": "0.00001", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1715558400, + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", + "id": "openai/gpt-4o", + "name": "OpenAI: GPT-4o", + "per_request_limits": null, + "pricing": { + "completion": "0.000015", + "image": "0.007225", + "prompt": "0.000005", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1715558400, + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", + "id": "openai/gpt-4o-2024-05-13", + "name": "OpenAI: GPT-4o (2024-05-13)", + "per_request_limits": null, + "pricing": { + "completion": "0.000015", + "image": "0.007225", + "prompt": "0.000005", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 4096 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1722902400, + "description": "The 2024-08-06 version of GPT-4o offers improved performance in structured outputs, with the ability to supply a JSON schema in the respone_format. Read more [here](https://openai.com/index/introducing-structured-outputs-in-the-api/).\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)", + "id": "openai/gpt-4o-2024-08-06", + "name": "OpenAI: GPT-4o (2024-08-06)", + "per_request_limits": null, + "pricing": { + "completion": "0.00001", + "image": "0.0036125", + "prompt": "0.0000025", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 16384 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1715558400, + "description": "GPT-4o Extended is an experimental variant of GPT-4o with an extended max output tokens. This model supports only text input to text output.\n\n_These are extended-context endpoints for [GPT-4o](/models/openai/gpt-4o). They may have higher prices._", + "id": "openai/gpt-4o:extended", + "name": "OpenAI: GPT-4o (extended)", + "per_request_limits": null, + "pricing": { + "completion": "0.000018", + "image": "0.007225", + "prompt": "0.000006", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": false, + "max_completion_tokens": 64000 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1721260800, + "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.", + "id": "openai/gpt-4o-mini", + "name": "OpenAI: GPT-4o-mini", + "per_request_limits": null, + "pricing": { + "completion": "0.0000006", + "image": "0.007225", + "prompt": "0.00000015", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 16384 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text+image->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1721260800, + "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.", + "id": "openai/gpt-4o-mini-2024-07-18", + "name": "OpenAI: GPT-4o-mini (2024-07-18)", + "per_request_limits": null, + "pricing": { + "completion": "0.0000006", + "image": "0.007225", + "prompt": "0.00000015", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 16384 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1726099200, + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "openai/o1-mini", + "name": "OpenAI: o1-mini", + "per_request_limits": null, + "pricing": { + "completion": "0.000012", + "image": "0", + "prompt": "0.000003", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 65536 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1726099200, + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "openai/o1-mini-2024-09-12", + "name": "OpenAI: o1-mini (2024-09-12)", + "per_request_limits": null, + "pricing": { + "completion": "0.000012", + "image": "0", + "prompt": "0.000003", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 65536 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1726099200, + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "openai/o1-preview", + "name": "OpenAI: o1-preview", + "per_request_limits": null, + "pricing": { + "completion": "0.00006", + "image": "0", + "prompt": "0.000015", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 32768 + } + }, + { + "architecture": { + "instruct_type": null, + "modality": "text->text", + "tokenizer": "GPT" + }, + "context_length": 128000, + "created": 1726099200, + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "id": "openai/o1-preview-2024-09-12", + "name": "OpenAI: o1-preview (2024-09-12)", + "per_request_limits": null, + "pricing": { + "completion": "0.00006", + "image": "0", + "prompt": "0.000015", + "request": "0" + }, + "top_provider": { + "context_length": 128000, + "is_moderated": true, + "max_completion_tokens": 32768 + } + }, + { + "architecture": { + "instruct_type": "openchat", + "modality": "text->text", + "tokenizer": "Mistral" + }, + "context_length": 8192, + "created": 1701129600, + "description": "OpenChat 7B is a library of open-source language models, fine-tuned with \"C-RLFT (Conditioned Reinforcement Learning Fine-Tuning)\" - a strategy inspired by offline reinforcement learning. It has been trained on mixed-quality data without preference labels.\n\n- For OpenChat fine-tuned on Mistral 7B, check out [OpenChat 7B](/models/openchat/openchat-7b).\n- For OpenChat fine-tuned on Llama 8B, check out [OpenChat 8B](/models/openchat/openchat-8b).\n\n#open-source", + "id": "openchat/openchat-7b", + "name": "OpenChat 3.5 7B", + "per_request_limits": null, + "pricing": { + "completion": "0.000000055", "image": "0", - "prompt": "0", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "openchat", "modality": "text->text", "tokenizer": "Mistral" }, - "context_length": 4096, - "created": 1699574400, - "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", - "id": "undi95/toppy-m-7b", - "name": "Toppy M 7B", + "context_length": 8192, + "created": 1701129600, + "description": "OpenChat 7B is a library of open-source language models, fine-tuned with \"C-RLFT (Conditioned Reinforcement Learning Fine-Tuning)\" - a strategy inspired by offline reinforcement learning. It has been trained on mixed-quality data without preference labels.\n\n- For OpenChat fine-tuned on Mistral 7B, check out [OpenChat 7B](/models/openchat/openchat-7b).\n- For OpenChat fine-tuned on Llama 8B, check out [OpenChat 8B](/models/openchat/openchat-8b).\n\n#open-source\n\n_These are free, rate-limited endpoints for [OpenChat 3.5 7B](/models/openchat/openchat-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "openchat/openchat-7b:free", + "name": "OpenChat 3.5 7B (free)", "per_request_limits": null, "pricing": { - "completion": "0.00000007", + "completion": "0", "image": "0", - "prompt": "0.00000007", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", "tokenizer": "Mistral" }, "context_length": 4096, - "created": 1699574400, - "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored\n\n_These are higher-throughput endpoints for [Toppy M 7B](/models/undi95/toppy-m-7b). They may have higher prices._", - "id": "undi95/toppy-m-7b:nitro", - "name": "Toppy M 7B (nitro)", + "created": 1700438400, + "description": "A continuation of [OpenHermes 2 model](/models/teknium/openhermes-2-mistral-7b), trained on additional code datasets.\nPotentially the most interesting finding from training on a good ratio (est. of around 7-14% of the total dataset) of code instruction was that it has boosted several non-code benchmarks, including TruthfulQA, AGIEval, and GPT4All suite. It did however reduce BigBench benchmark score, but the net gain overall is significant.", + "id": "teknium/openhermes-2.5-mistral-7b", + "name": "OpenHermes 2.5 Mistral 7B", "per_request_limits": null, "pricing": { - "completion": "0.00000007", + "completion": "0.00000017", "image": "0", - "prompt": "0.00000007", + "prompt": "0.00000017", "request": "0" }, "top_provider": { @@ -3268,22 +3700,22 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "Router" + "tokenizer": "Llama3" }, - "context_length": 200000, - "created": 1699401600, - "description": "Depending on their size, subject, and complexity, your prompts will be sent to [Llama 3 70B Instruct](/models/meta-llama/llama-3-70b-instruct), [Claude 3.5 Sonnet (self-moderated)](/models/anthropic/claude-3.5-sonnet:beta) or [GPT-4o](/models/openai/gpt-4o). To see which model was used, visit [Activity](/activity).\n\nA major redesign of this router is coming soon. Stay tuned on [Discord](https://discord.gg/fVyRaUDgxW) for updates.", - "id": "openrouter/auto", - "name": "Auto (best for prompt)", + "context_length": 127072, + "created": 1723593600, + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance. The model is built upon the Llama 3.1 405B and has internet access.", + "id": "perplexity/llama-3.1-sonar-huge-128k-online", + "name": "Perplexity: Llama 3.1 Sonar 405B Online", "per_request_limits": null, "pricing": { - "completion": "-1", - "image": "-1", - "prompt": "-1", - "request": "-1" + "completion": "0.000005", + "image": "0", + "prompt": "0.000005", + "request": "0.005" }, "top_provider": { - "context_length": null, + "context_length": 127072, "is_moderated": false, "max_completion_tokens": null } @@ -3292,528 +3724,528 @@ "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama3" }, - "context_length": 128000, - "created": 1699228800, - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", - "id": "openai/gpt-4-1106-preview", - "name": "OpenAI: GPT-4 Turbo (older v1106)", + "context_length": 131072, + "created": 1722470400, + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3.1-sonar-large-128k-online) of this model has Internet access.", + "id": "perplexity/llama-3.1-sonar-large-128k-chat", + "name": "Perplexity: Llama 3.1 Sonar 70B", "per_request_limits": null, "pricing": { - "completion": "0.00003", + "completion": "0.000001", "image": "0", - "prompt": "0.00001", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 128000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 131072, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama3" }, - "context_length": 16385, - "created": 1699228800, - "description": "An older GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.", - "id": "openai/gpt-3.5-turbo-1106", - "name": "OpenAI: GPT-3.5 Turbo 16k (older v1106)", + "context_length": 127072, + "created": 1722470400, + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "id": "perplexity/llama-3.1-sonar-large-128k-online", + "name": "Perplexity: Llama 3.1 Sonar 70B Online", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.000001", "image": "0", "prompt": "0.000001", - "request": "0" + "request": "0.005" }, "top_provider": { - "context_length": 16385, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 127072, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "PaLM" + "tokenizer": "Llama3" }, - "context_length": 131040, - "created": 1698969600, - "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", - "id": "google/palm-2-codechat-bison-32k", - "name": "Google: PaLM 2 Code Chat 32k", + "context_length": 131072, + "created": 1722470400, + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3.1-sonar-small-128k-online) of this model has Internet access.", + "id": "perplexity/llama-3.1-sonar-small-128k-chat", + "name": "Perplexity: Llama 3.1 Sonar 8B", "per_request_limits": null, "pricing": { - "completion": "0.0000005", + "completion": "0.0000002", "image": "0", - "prompt": "0.00000025", + "prompt": "0.0000002", "request": "0" }, "top_provider": { "context_length": 131072, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": null } }, { "architecture": { "instruct_type": null, "modality": "text->text", - "tokenizer": "PaLM" + "tokenizer": "Llama3" }, - "context_length": 131040, - "created": 1698969600, - "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", - "id": "google/palm-2-chat-bison-32k", - "name": "Google: PaLM 2 Chat 32k", + "context_length": 127072, + "created": 1722470400, + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "id": "perplexity/llama-3.1-sonar-small-128k-online", + "name": "Perplexity: Llama 3.1 Sonar 8B Online", "per_request_limits": null, "pricing": { - "completion": "0.0000005", + "completion": "0.0000002", "image": "0", - "prompt": "0.00000025", - "request": "0" + "prompt": "0.0000002", + "request": "0.005" }, "top_provider": { - "context_length": 131072, + "context_length": 127072, "is_moderated": false, - "max_completion_tokens": 32768 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "airoboros", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Llama3" }, - "context_length": 4096, - "created": 1698537600, - "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", - "id": "jondurbin/airoboros-l2-70b", - "name": "Airoboros 70B", + "context_length": 32768, + "created": 1715644800, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-large-32k-online) of this model has Internet access.", + "id": "perplexity/llama-3-sonar-large-32k-chat", + "name": "Perplexity: Llama3 Sonar 70B", "per_request_limits": null, "pricing": { - "completion": "0.0000005", + "completion": "0.000001", "image": "0", - "prompt": "0.0000005", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "airoboros", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Llama3" }, - "context_length": 8192, - "created": 1697328000, - "description": "Xwin-LM aims to develop and open-source alignment tech for LLMs. Our first release, built-upon on the [Llama2](/models/${Model.Llama_2_13B_Chat}) base models, ranked TOP-1 on AlpacaEval. Notably, it's the first to surpass [GPT-4](/models/${Model.GPT_4}) on this benchmark. The project will be continuously updated.", - "id": "xwin-lm/xwin-lm-70b", - "name": "Xwin 70B", + "context_length": 28000, + "created": 1715644800, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-large-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "id": "perplexity/llama-3-sonar-large-32k-online", + "name": "Perplexity: Llama3 Sonar 70B Online", "per_request_limits": null, "pricing": { - "completion": "0.00000375", + "completion": "0.000001", "image": "0", - "prompt": "0.00000375", - "request": "0" + "prompt": "0.000001", + "request": "0.005" }, "top_provider": { - "context_length": 8192, + "context_length": 28000, "is_moderated": false, - "max_completion_tokens": 400 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "mistral", + "instruct_type": null, "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama3" }, - "context_length": 4096, - "created": 1695859200, - "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", - "id": "mistralai/mistral-7b-instruct-v0.1", - "name": "Mistral: Mistral 7B Instruct v0.1", + "context_length": 32768, + "created": 1715644800, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-small-32k-online) of this model has Internet access.", + "id": "perplexity/llama-3-sonar-small-32k-chat", + "name": "Perplexity: Llama3 Sonar 8B", "per_request_limits": null, "pricing": { - "completion": "0.000000055", + "completion": "0.0000002", "image": "0", - "prompt": "0.000000055", + "prompt": "0.0000002", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 32768, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "chatml", + "instruct_type": null, "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama3" }, - "context_length": 4095, - "created": 1695859200, - "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", - "id": "openai/gpt-3.5-turbo-instruct", - "name": "OpenAI: GPT-3.5 Turbo Instruct", + "context_length": 28000, + "created": 1715644800, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-small-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "id": "perplexity/llama-3-sonar-small-32k-online", + "name": "Perplexity: Llama3 Sonar 8B Online", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.0000002", "image": "0", - "prompt": "0.0000015", - "request": "0" + "prompt": "0.0000002", + "request": "0.005" }, "top_provider": { - "context_length": 4095, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 28000, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "phi3", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Other" }, - "context_length": 8192, - "created": 1693612800, - "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", - "id": "pygmalionai/mythalion-13b", - "name": "Pygmalion: Mythalion 13B", + "context_length": 128000, + "created": 1716508800, + "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", + "id": "microsoft/phi-3-medium-128k-instruct", + "name": "Phi-3 Medium 128K Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000001125", + "completion": "0.000001", "image": "0", - "prompt": "0.000001125", + "prompt": "0.000001", "request": "0" }, "top_provider": { - "context_length": 8192, + "context_length": 128000, "is_moderated": false, - "max_completion_tokens": 400 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "phi3", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 32767, - "created": 1693180800, - "description": "GPT-4-32k is an extended version of GPT-4, with the same capabilities but quadrupled context length, allowing for processing up to 40 pages of text in a single pass. This is particularly beneficial for handling longer content like interacting with PDFs without an external vector database. Training data: up to Sep 2021.", - "id": "openai/gpt-4-32k-0314", - "name": "OpenAI: GPT-4 32k (older v0314)", + "context_length": 128000, + "created": 1716508800, + "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).\n\n_These are free, rate-limited endpoints for [Phi-3 Medium 128K Instruct](/models/microsoft/phi-3-medium-128k-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "microsoft/phi-3-medium-128k-instruct:free", + "name": "Phi-3 Medium 128K Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.00012", + "completion": "0", "image": "0", - "prompt": "0.00006", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 32767, - "is_moderated": true, + "context_length": 8192, + "is_moderated": false, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, + "instruct_type": "phi3", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 32767, - "created": 1693180800, - "description": "GPT-4-32k is an extended version of GPT-4, with the same capabilities but quadrupled context length, allowing for processing up to 40 pages of text in a single pass. This is particularly beneficial for handling longer content like interacting with PDFs without an external vector database. Training data: up to Sep 2021.", - "id": "openai/gpt-4-32k", - "name": "OpenAI: GPT-4 32k", + "context_length": 128000, + "created": 1716681600, + "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", + "id": "microsoft/phi-3-mini-128k-instruct", + "name": "Phi-3 Mini 128K Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00012", + "completion": "0.0000001", "image": "0", - "prompt": "0.00006", + "prompt": "0.0000001", "request": "0" }, "top_provider": { - "context_length": 32767, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 128000, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "phi3", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Other" }, - "context_length": 16385, - "created": 1693180800, - "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", - "id": "openai/gpt-3.5-turbo-16k", - "name": "OpenAI: GPT-3.5 Turbo 16k", + "context_length": 128000, + "created": 1716681600, + "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.\n\n_These are free, rate-limited endpoints for [Phi-3 Mini 128K Instruct](/models/microsoft/phi-3-mini-128k-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "microsoft/phi-3-mini-128k-instruct:free", + "name": "Phi-3 Mini 128K Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.000004", + "completion": "0", "image": "0", - "prompt": "0.000003", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 16385, - "is_moderated": true, + "context_length": 8192, + "is_moderated": false, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "phi3", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Other" }, - "context_length": 4096, - "created": 1692489600, - "description": "A state-of-the-art language model fine-tuned on over 300k instructions by Nous Research, with Teknium and Emozilla leading the fine tuning process.", - "id": "nousresearch/nous-hermes-llama2-13b", - "name": "Nous: Hermes 13B", + "context_length": 128000, + "created": 1724198400, + "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", + "id": "microsoft/phi-3.5-mini-128k-instruct", + "name": "Phi-3.5 Mini 128K Instruct", "per_request_limits": null, "pricing": { - "completion": "0.00000017", + "completion": "0.0000001", "image": "0", - "prompt": "0.00000017", + "prompt": "0.0000001", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 128000, "is_moderated": false, "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "zephyr", + "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Mistral" + "tokenizer": "Llama2" }, - "context_length": 4096, - "created": 1690934400, - "description": "Zephyr is a series of language models that are trained to act as helpful assistants. Zephyr-7B-β is the second model in the series, and is a fine-tuned version of [mistralai/Mistral-7B-v0.1](/models/mistralai/mistral-7b-instruct-v0.1) that was trained on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO).\n\n_These are free, rate-limited endpoints for [Zephyr 7B](/models/huggingfaceh4/zephyr-7b-beta). Outputs may be cached. Read about rate limits [here](/docs/limits)._", - "id": "huggingfaceh4/zephyr-7b-beta:free", - "name": "Hugging Face: Zephyr 7B (free)", + "context_length": 8192, + "created": 1693612800, + "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", + "id": "pygmalionai/mythalion-13b", + "name": "Pygmalion: Mythalion 13B", "per_request_limits": null, "pricing": { - "completion": "0", + "completion": "0.000001125", "image": "0", - "prompt": "0", + "prompt": "0.000001125", "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 8192, "is_moderated": false, - "max_completion_tokens": 2048 + "max_completion_tokens": 400 } }, { "architecture": { - "instruct_type": "alpaca", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Qwen" }, - "context_length": 8000, - "created": 1690934400, - "description": "An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.", - "id": "mancer/weaver", - "name": "Mancer: Weaver (alpha)", + "context_length": 32768, + "created": 1715212800, + "description": "Qwen1.5 110B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-110b-chat", + "name": "Qwen 1.5 110B Chat", "per_request_limits": null, "pricing": { - "completion": "0.00000225", + "completion": "0.00000162", "image": "0", - "prompt": "0.000001875", + "prompt": "0.00000162", "request": "0" }, "top_provider": { - "context_length": 8000, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": 1000 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "claude", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", - "id": "anthropic/claude-instant-1.0", - "name": "Anthropic: Claude Instant v1.0", + "context_length": 32768, + "created": 1715212800, + "description": "Qwen1.5 72B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-72b-chat", + "name": "Qwen 1.5 72B Chat", "per_request_limits": null, "pricing": { - "completion": "0.0000024", + "completion": "0.00000081", "image": "0", - "prompt": "0.0000008", + "prompt": "0.00000081", "request": "0" }, "top_provider": { - "context_length": 100000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "claude", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", - "id": "anthropic/claude-1.2", - "name": "Anthropic: Claude v1.2", + "context_length": 32768, + "created": 1717718400, + "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-2-72b-instruct", + "name": "Qwen 2 72B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.00000039", "image": "0", - "prompt": "0.000008", + "prompt": "0.00000034", "request": "0" }, "top_provider": { - "context_length": 100000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": "claude", + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", - "id": "anthropic/claude-1", - "name": "Anthropic: Claude v1", + "context_length": 32768, + "created": 1721088000, + "description": "Qwen2 7B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-2-7b-instruct", + "name": "Qwen 2 7B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.000000054", "image": "0", - "prompt": "0.000008", + "prompt": "0.000000054", "request": "0" }, "top_provider": { - "context_length": 100000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32768, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "Claude" + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", - "id": "anthropic/claude-instant-1", - "name": "Anthropic: Claude Instant v1", + "context_length": 32768, + "created": 1721088000, + "description": "Qwen2 7B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).\n\n_These are free, rate-limited endpoints for [Qwen 2 7B Instruct](/models/qwen/qwen-2-7b-instruct). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "qwen/qwen-2-7b-instruct:free", + "name": "Qwen 2 7B Instruct (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000024", + "completion": "0", "image": "0", - "prompt": "0.0000008", + "prompt": "0", "request": "0" }, "top_provider": { - "context_length": 100000, - "is_moderated": true, + "context_length": 8192, + "is_moderated": false, "max_completion_tokens": 4096 } }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "Claude" + "instruct_type": "chatml", + "modality": "text+image->text", + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-instant-1) variant._", - "id": "anthropic/claude-instant-1:beta", - "name": "Anthropic: Claude Instant v1 (self-moderated)", + "context_length": 32768, + "created": 1726617600, + "description": "Qwen2 VL 72B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-2-vl-72b-instruct", + "name": "Qwen2-VL 72B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.0000024", - "image": "0", - "prompt": "0.0000008", + "completion": "0.0000004", + "image": "0.000578", + "prompt": "0.0000004", "request": "0" }, "top_provider": { - "context_length": 100000, + "context_length": 4096, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "Claude" - }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", - "id": "anthropic/claude-2.0", - "name": "Anthropic: Claude v2.0", + "instruct_type": "chatml", + "modality": "text+image->text", + "tokenizer": "Qwen" + }, + "context_length": 32768, + "created": 1724803200, + "description": "Qwen2 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-2-vl-7b-instruct", + "name": "Qwen2-VL 7B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000024", - "image": "0", - "prompt": "0.000008", + "completion": "0.0000001", + "image": "0.0001445", + "prompt": "0.0000001", "request": "0" }, "top_provider": { - "context_length": 100000, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 4096, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "Claude" + "instruct_type": "chatml", + "modality": "text+image->text", + "tokenizer": "Qwen" }, - "context_length": 100000, - "created": 1690502400, - "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/models/anthropic/claude-2.0) variant._", - "id": "anthropic/claude-2.0:beta", - "name": "Anthropic: Claude v2.0 (self-moderated)", + "context_length": 131072, + "created": 1726704000, + "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "id": "qwen/qwen-2.5-72b-instruct", + "name": "Qwen2.5 72B Instruct", "per_request_limits": null, "pricing": { - "completion": "0.000024", + "completion": "0.0000004", "image": "0", - "prompt": "0.000008", + "prompt": "0.00000035", "request": "0" }, "top_provider": { - "context_length": 100000, + "context_length": 32000, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { @@ -3835,7 +4267,7 @@ "request": "0" }, "top_provider": { - "context_length": 4096, + "context_length": 6144, "is_moderated": false, "max_completion_tokens": 400 } @@ -3866,39 +4298,15 @@ }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "PaLM" - }, - "context_length": 28672, - "created": 1689811200, - "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", - "id": "google/palm-2-codechat-bison", - "name": "Google: PaLM 2 Code Chat", - "per_request_limits": null, - "pricing": { - "completion": "0.0000005", - "image": "0", - "prompt": "0.00000025", - "request": "0" - }, - "top_provider": { - "context_length": 28672, - "is_moderated": false, - "max_completion_tokens": 4096 - } - }, - { - "architecture": { - "instruct_type": null, + "instruct_type": "chatml", "modality": "text->text", - "tokenizer": "PaLM" + "tokenizer": "Qwen" }, - "context_length": 36864, - "created": 1689811200, - "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", - "id": "google/palm-2-chat-bison", - "name": "Google: PaLM 2 Chat", + "context_length": 32768, + "created": 1727654400, + "description": "Rocinante 12B is designed for engaging storytelling and rich prose.\n\nEarly testers have reported:\n- Expanded vocabulary with unique and expressive word choices\n- Enhanced creativity for vivid narratives\n- Adventure-filled and captivating stories", + "id": "thedrummer/rocinante-12b", + "name": "Rocinante 12B", "per_request_limits": null, "pricing": { "completion": "0.0000005", @@ -3907,27 +4315,27 @@ "request": "0" }, "top_provider": { - "context_length": 36864, + "context_length": 32768, "is_moderated": false, - "max_completion_tokens": 4096 + "max_completion_tokens": null } }, { "architecture": { "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Mistral" }, "context_length": 4096, - "created": 1688256000, - "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", - "id": "gryphe/mythomax-l2-13b", - "name": "MythoMax 13B", + "created": 1699574400, + "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", + "id": "undi95/toppy-m-7b", + "name": "Toppy M 7B", "per_request_limits": null, "pricing": { - "completion": "0.0000001", + "completion": "0.00000007", "image": "0", - "prompt": "0.0000001", + "prompt": "0.00000007", "request": "0" }, "top_provider": { @@ -3940,66 +4348,42 @@ "architecture": { "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Mistral" }, "context_length": 4096, - "created": 1688256000, - "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge\n\n_These are higher-throughput endpoints for [MythoMax 13B](/models/gryphe/mythomax-l2-13b). They may have higher prices._", - "id": "gryphe/mythomax-l2-13b:nitro", - "name": "MythoMax 13B (nitro)", + "created": 1699574400, + "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored\n\n_These are free, rate-limited endpoints for [Toppy M 7B](/models/undi95/toppy-m-7b). Outputs may be cached. Read about rate limits [here](/docs/limits)._", + "id": "undi95/toppy-m-7b:free", + "name": "Toppy M 7B (free)", "per_request_limits": null, "pricing": { - "completion": "0.0000002", + "completion": "0", "image": "0", - "prompt": "0.0000002", + "prompt": "0", "request": "0" }, "top_provider": { "context_length": 4096, "is_moderated": false, - "max_completion_tokens": null + "max_completion_tokens": 2048 } }, { "architecture": { "instruct_type": "alpaca", "modality": "text->text", - "tokenizer": "Llama2" - }, - "context_length": 8192, - "created": 1688256000, - "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge\n\n_These are extended-context endpoints for [MythoMax 13B](/models/gryphe/mythomax-l2-13b). They may have higher prices._", - "id": "gryphe/mythomax-l2-13b:extended", - "name": "MythoMax 13B (extended)", - "per_request_limits": null, - "pricing": { - "completion": "0.000001125", - "image": "0", - "prompt": "0.000001125", - "request": "0" - }, - "top_provider": { - "context_length": 8192, - "is_moderated": false, - "max_completion_tokens": 400 - } - }, - { - "architecture": { - "instruct_type": "llama2", - "modality": "text->text", - "tokenizer": "Llama2" + "tokenizer": "Mistral" }, "context_length": 4096, - "created": 1687219200, - "description": "A 13 billion parameter language model from Meta, fine tuned for chat completions", - "id": "meta-llama/llama-2-13b-chat", - "name": "Meta: Llama v2 13B Chat", + "created": 1699574400, + "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored\n\n_These are higher-throughput endpoints for [Toppy M 7B](/models/undi95/toppy-m-7b). They may have higher prices._", + "id": "undi95/toppy-m-7b:nitro", + "name": "Toppy M 7B (nitro)", "per_request_limits": null, "pricing": { - "completion": "0.00000027", + "completion": "0.00000007", "image": "0", - "prompt": "0.00000027", + "prompt": "0.00000007", "request": "0" }, "top_provider": { @@ -4010,122 +4394,74 @@ }, { "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "GPT" - }, - "context_length": 8191, - "created": 1685232000, - "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", - "id": "openai/gpt-4-0314", - "name": "OpenAI: GPT-4 (older v0314)", - "per_request_limits": null, - "pricing": { - "completion": "0.00006", - "image": "0", - "prompt": "0.00003", - "request": "0" - }, - "top_provider": { - "context_length": 8191, - "is_moderated": true, - "max_completion_tokens": 4096 - } - }, - { - "architecture": { - "instruct_type": null, - "modality": "text->text", - "tokenizer": "GPT" - }, - "context_length": 8191, - "created": 1685232000, - "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", - "id": "openai/gpt-4", - "name": "OpenAI: GPT-4", - "per_request_limits": null, - "pricing": { - "completion": "0.00006", - "image": "0", - "prompt": "0.00003", - "request": "0" - }, - "top_provider": { - "context_length": 8191, - "is_moderated": true, - "max_completion_tokens": 4096 - } - }, - { - "architecture": { - "instruct_type": null, + "instruct_type": "vicuna", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Mistral" }, - "context_length": 4095, - "created": 1685232000, - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", - "id": "openai/gpt-3.5-turbo-0301", - "name": "OpenAI: GPT-3.5 Turbo (older v0301)", + "context_length": 32000, + "created": 1713225600, + "description": "WizardLM-2 7B is the smaller variant of Microsoft AI's latest Wizard model. It is the fastest and achieves comparable performance with existing 10x larger opensource leading models\n\nIt is a finetune of [Mistral 7B Instruct](/models/mistralai/mistral-7b-instruct), using the same technique as [WizardLM-2 8x22B](/models/microsoft/wizardlm-2-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", + "id": "microsoft/wizardlm-2-7b", + "name": "WizardLM-2 7B", "per_request_limits": null, "pricing": { - "completion": "0.000002", + "completion": "0.000000055", "image": "0", - "prompt": "0.000001", + "prompt": "0.000000055", "request": "0" }, "top_provider": { - "context_length": 4095, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 32000, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "vicuna", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Mistral" }, - "context_length": 16385, - "created": 1685232000, - "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", - "id": "openai/gpt-3.5-turbo-0125", - "name": "OpenAI: GPT-3.5 Turbo 16k", + "context_length": 65536, + "created": 1713225600, + "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", + "id": "microsoft/wizardlm-2-8x22b", + "name": "WizardLM-2 8x22B", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0.0000005", "image": "0", "prompt": "0.0000005", "request": "0" }, "top_provider": { - "context_length": 16385, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 65536, + "is_moderated": false, + "max_completion_tokens": null } }, { "architecture": { - "instruct_type": null, + "instruct_type": "airoboros", "modality": "text->text", - "tokenizer": "GPT" + "tokenizer": "Llama2" }, - "context_length": 16385, - "created": 1685232000, - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", - "id": "openai/gpt-3.5-turbo", - "name": "OpenAI: GPT-3.5 Turbo", + "context_length": 8192, + "created": 1697328000, + "description": "Xwin-LM aims to develop and open-source alignment tech for LLMs. Our first release, built-upon on the [Llama2](/models/${Model.Llama_2_13B_Chat}) base models, ranked TOP-1 on AlpacaEval. Notably, it's the first to surpass [GPT-4](/models/${Model.GPT_4}) on this benchmark. The project will be continuously updated.", + "id": "xwin-lm/xwin-lm-70b", + "name": "Xwin 70B", "per_request_limits": null, "pricing": { - "completion": "0.0000015", + "completion": "0.00000375", "image": "0", - "prompt": "0.0000005", + "prompt": "0.00000375", "request": "0" }, "top_provider": { - "context_length": 16385, - "is_moderated": true, - "max_completion_tokens": 4096 + "context_length": 8192, + "is_moderated": false, + "max_completion_tokens": 400 } } ] diff --git a/sortModelsJsonKeys.js b/sortModelsJsonKeys.js index f332a5746..2c1b75206 100644 --- a/sortModelsJsonKeys.js +++ b/sortModelsJsonKeys.js @@ -3,6 +3,11 @@ import fs from 'fs/promises'; // Function to recursively sort object keys function sortObjectKeys(obj) { if (Array.isArray(obj)) { + // Check if the array contains objects with a 'name' field + if (obj.length > 0 && typeof obj[0] === 'object' && 'name' in obj[0]) { + // Sort the array by the 'name' field + obj.sort((a, b) => a.name.localeCompare(b.name)); + } return obj.map(sortObjectKeys); } else if (obj !== null && typeof obj === 'object') { return Object.keys(obj).sort().reduce((sortedObj, key) => {