README_EN.md

简体中文

Note

Please note: The experience key is public, but please do not abuse it. If you need to modify the code for self-deployment of this project, please modify the default experience key at the same time, or at least do not remove the related restrictions.

Silo - Pure front-end multi-model chat, text-to-image generation

Silo is an application mainly based on the SiliconCloud API, which currently supports multiple large models for simultaneous chat、text to image, with extremely fast response times. It supports model parameter adjustments and automatic balance refresh functions.

You can also conveniently add models such as Google Gemini, Claude, X-AI/Grok, OpenAI, and others.

And it does not require a backend/server.

Features

🌐 Browser extension and web access available, and you can chat with page by using extension
🚀 Based on SiliconCloud API, rich models and fast response
🔑 Simple configuration, you only need to configure the key once
💰 Support SiliconCloud balance automatic refresh
🧩 Provides the ability to add custom models to use models outside of the API。Quickly supports Gemini, Claude, DeepSeek, ZhiPu, and provides OpenAI-compatible configurations, as well as a hardcore method of writing your own code.
🔄 Support adjusting model order
🌙 Night Mode
🧘‍♂️ Zen mode
🇬🇧 i18n support
📱 Mobile phone support (limited by screen size, a simpler version is provided)
📦 Pure front-end implementation (convenient for migration, CDN deployment, etc., does not rely on server transfer)
🔧 Model parameter adjustment
💬 Optimized input experience

Browser Extension

When the version of this project changes, it will be automatically packaged and submitted to the Chrome App Store and Edge Addons, and the generated zip file will be uploaded to GitHub Releases. You can also download it directly and install it in the browser.

Docker deployment

version: '3.8'

services:
  frontend:
    image: ghcr.io/KwokKwok/silo:latest
    ports:
      - '3000:3000'
    environment:
      # Set the default experience SiliconFlow API key.
      - SILO_EXPERIENCE_SK=
      # Default activated chat model
      - SILO_DEFAULT_ACTIVE_CHAT_MODELS=Qwen/Qwen2.5-7B-Instruct,THUDM/glm-4-9b-chat,01-ai/Yi-1.5-9B-Chat-16K

Environment Variables

Note

This section is not of concern to regular users, only those who are deploying the project themselves may need to understand it. Environment variables configured may need to be restarted or re-deployed. Please verify whether it is effective.

Note

As a reminder, this project is purely frontend-based, and the configured keys may be maliciously used, so please configure with caution.

- SILO_EXPERIENCE_SK: SiliconFlow experience API key; you can enter '0' indicate not providing this option
- SILO_PAID_SK: SiliconFlow paid API key; when set, there will be no experience key notification or restrictions
- PAID_SK_PASSWORD: Password to protect the paid key, after setting, users can automatically use the paid key by entering the password in the user interface. It is strongly recommended to set this variable at the same time as SILO_PAID_SK. Please note that this variable does not start with SILO_
- SILO_AFF_LINK: Registration link in the SiliconFlow key popup
- SILO_DEFAULT_ACTIVE_CHAT_MODELS: Default activated chat models, multiple models separated by commas
- SILO_DEFAULT_ACTIVE_IMAGE_MODELS: Default activated text-to-image models, multiple models separated by commas

Acknowledgments

1. Thanks to SiliconCloud.
2. The chat page was referenced from Vercel AI Playground, and the column division was referenced from ChatHub.
3. The translation used the default prompt of Immersive Translate, which was very effective.

License

This project is licensed under the MIT License. See LICENSE for more details.

More

Buy Me A Coffee

Buy me a coffee

Thanks to the suggestion from 黄少侠@Jike. This project is open source and completely free. If you find this project useful, feel free to Buy me a coffee~