Merge remote-tracking branch 'upstream/main'

.env.example

@@ -65,6 +65,7 @@ PROXY=
 # ANYSCALE_API_KEY=
 # APIPIE_API_KEY=
 # COHERE_API_KEY=
+# DEEPSEEK_API_KEY=
 # DATABRICKS_API_KEY=
 # FIREWORKS_API_KEY=
 # GROQ_API_KEY=
@@ -74,6 +75,7 @@ PROXY=
 # PERPLEXITY_API_KEY=
 # SHUTTLEAI_API_KEY=
 # TOGETHERAI_API_KEY=
+# UNIFY_API_KEY=
 
 #============#
 # Anthropic  #
@@ -429,10 +431,10 @@ ALLOW_SHARED_LINKS_PUBLIC=true
 # Static File Cache Control    #
 #==============================#
 
-# Leave commented out to use default of 1 month for max-age and 1 week for s-maxage
+# Leave commented out to use defaults: 1 day (86400 seconds) for s-maxage and 2 days (172800 seconds) for max-age
 # NODE_ENV must be set to production for these to take effect
-# STATIC_CACHE_MAX_AGE=604800
-# STATIC_CACHE_S_MAX_AGE=259200
+# STATIC_CACHE_MAX_AGE=172800
+# STATIC_CACHE_S_MAX_AGE=86400
 
 # If you have another service in front of your LibreChat doing compression, disable express based compression here
 # DISABLE_COMPRESSION=true

.github/workflows/frontend-review.yml

@@ -53,4 +53,4 @@ jobs:
 
       - name: Run unit tests
         run: npm run test:ci --verbose
-        working-directory: client
+        working-directory: client

Dockerfile

@@ -1,4 +1,4 @@
-# v0.7.4
+# v0.7.5-rc1
 
 # Base node image
 FROM node:20-alpine AS node

Dockerfile.multi

@@ -1,43 +1,44 @@
-# v0.7.4
+# Dockerfile.multi
+# v0.7.5-rc1
 
-# Build API, Client and Data Provider
+# Base for all builds
 FROM node:20-alpine AS base
+WORKDIR /app
+RUN apk --no-cache add curl
+RUN npm config set fetch-retry-maxtimeout 600000 && \
+    npm config set fetch-retries 5 && \
+    npm config set fetch-retry-mintimeout 15000
+COPY package*.json ./
+COPY packages/data-provider/package*.json ./packages/data-provider/
+COPY client/package*.json ./client/
+COPY api/package*.json ./api/
+RUN npm ci
 
 # Build data-provider
 FROM base AS data-provider-build
 WORKDIR /app/packages/data-provider
-COPY ./packages/data-provider ./
-RUN npm install; npm cache clean --force
+COPY packages/data-provider ./
 RUN npm run build
 RUN npm prune --production
 
-# React client build
+# Client build
 FROM base AS client-build
 WORKDIR /app/client
-COPY ./client/package*.json ./
-# Copy data-provider to client's node_modules
-COPY --from=data-provider-build /app/packages/data-provider/ /app/client/node_modules/librechat-data-provider/
-RUN npm install; npm cache clean --force
-COPY ./client/ ./
+COPY client ./
+COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
 ENV NODE_OPTIONS="--max-old-space-size=2048"
 RUN npm run build
+RUN npm prune --production
 
-# Node API setup
+# API setup (including client dist)
 FROM base AS api-build
+WORKDIR /app
+COPY api ./api
+COPY config ./config
+COPY --from=data-provider-build /app/packages/data-provider/dist ./packages/data-provider/dist
+COPY --from=client-build /app/client/dist ./client/dist
 WORKDIR /app/api
-COPY api/package*.json ./
-COPY api/ ./
-# Copy helper scripts
-COPY config/ ./
-# Copy data-provider to API's node_modules
-COPY --from=data-provider-build /app/packages/data-provider/ /app/api/node_modules/librechat-data-provider/
-RUN npm install --include prod; npm cache clean --force
-COPY --from=client-build /app/client/dist /app/client/dist
+RUN npm prune --production
 EXPOSE 3080
 ENV HOST=0.0.0.0
 CMD ["node", "server/index.js"]
-
-# Nginx setup
-FROM nginx:1.27.0-alpine AS prod-stage
-COPY ./client/nginx.conf /etc/nginx/conf.d/default.conf
-CMD ["nginx", "-g", "daemon off;"]

README.md

@@ -45,6 +45,8 @@
   - OpenAI, Azure OpenAI, BingAI, ChatGPT, Google Vertex AI, Anthropic (Claude), Plugins, Assistants API (including Azure Assistants)
 - ✅ Compatible across both **[Remote & Local AI services](https://www.librechat.ai/docs/configuration/librechat_yaml/ai_endpoints):**
   - groq, Ollama, Cohere, Mistral AI, Apple MLX, koboldcpp, OpenRouter, together.ai, Perplexity, ShuttleAI, and more
+- 🪄 Generative UI with [Code Artifacts](https://youtu.be/GfTj7O4gmd0?si=WJbdnemZpJzBrJo3)
+   - Create React, HTML code, and Mermaid diagrams right in chat
 - 💾 Create, Save, & Share Custom Presets
 - 🔀 Switch between AI Endpoints and Presets, mid-chat
 - 🔄 Edit, Resubmit, and Continue Messages with Conversation branching

api/app/clients/AnthropicClient.js

@@ -83,16 +83,19 @@ class AnthropicClient extends BaseClient {
       this.options = options;
     }
 
-    const modelOptions = this.options.modelOptions || {};
-    this.modelOptions = {
-      ...modelOptions,
-      model: modelOptions.model || anthropicSettings.model.default,
-    };
+    this.modelOptions = Object.assign(
+      {
+        model: anthropicSettings.model.default,
+      },
+      this.modelOptions,
+      this.options.modelOptions,
+    );
 
     const modelMatch = matchModelName(this.modelOptions.model, EModelEndpoint.anthropic);
     this.isClaude3 = modelMatch.startsWith('claude-3');
     this.isLegacyOutput = !modelMatch.startsWith('claude-3-5-sonnet');
-    this.supportsCacheControl = this.checkPromptCacheSupport(modelMatch);
+    this.supportsCacheControl =
+      this.options.promptCache && this.checkPromptCacheSupport(modelMatch);
 
     if (
       this.isLegacyOutput &&
@@ -489,7 +492,10 @@ class AnthropicClient extends BaseClient {
       identityPrefix = `${identityPrefix}\nYou are ${this.options.modelLabel}`;
     }
 
-    let promptPrefix = (this.options.promptPrefix || '').trim();
+    let promptPrefix = (this.options.promptPrefix ?? '').trim();
+    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+      promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
+    }
     if (promptPrefix) {
       // If the prompt prefix doesn't end with the end token, add it.
       if (!promptPrefix.endsWith(`${this.endToken}`)) {
@@ -817,8 +823,10 @@ class AnthropicClient extends BaseClient {
   getSaveOptions() {
     return {
       maxContextTokens: this.options.maxContextTokens,
+      artifacts: this.options.artifacts,
       promptPrefix: this.options.promptPrefix,
       modelLabel: this.options.modelLabel,
+      promptCache: this.options.promptCache,
       resendFiles: this.options.resendFiles,
       iconURL: this.options.iconURL,
       greeting: this.options.greeting,

api/app/clients/BaseClient.js

@@ -1,6 +1,12 @@
 const crypto = require('crypto');
 const fetch = require('node-fetch');
-const { supportsBalanceCheck, Constants, CacheKeys, Time } = require('librechat-data-provider');
+const {
+  supportsBalanceCheck,
+  ErrorTypes,
+  Constants,
+  CacheKeys,
+  Time,
+} = require('librechat-data-provider');
 const { getMessages, saveMessage, updateMessage, saveConvo } = require('~/models');
 const { addSpaceIfNeeded, isEnabled } = require('~/server/utils');
 const checkBalance = require('~/models/checkBalance');
@@ -28,6 +34,12 @@ class BaseClient {
     this.userMessagePromise;
     /** @type {ClientDatabaseSavePromise} */
     this.responsePromise;
+    /** @type {string} */
+    this.user;
+    /** @type {string} */
+    this.conversationId;
+    /** @type {string} */
+    this.responseMessageId;
   }
 
   setOptions() {
@@ -155,6 +167,8 @@ class BaseClient {
       this.currentMessages[this.currentMessages.length - 1].messageId = head;
     }
 
+    this.responseMessageId = responseMessageId;
+
     return {
       ...opts,
       user,
@@ -341,7 +355,12 @@ class BaseClient {
     };
   }
 
-  async handleContextStrategy({ instructions, orderedMessages, formattedMessages }) {
+  async handleContextStrategy({
+    instructions,
+    orderedMessages,
+    formattedMessages,
+    buildTokenMap = true,
+  }) {
     let _instructions;
     let tokenCount;
 
@@ -383,9 +402,10 @@ class BaseClient {
 
     const latestMessage = orderedWithInstructions[orderedWithInstructions.length - 1];
     if (payload.length === 0 && !shouldSummarize && latestMessage) {
-      throw new Error(
-        `Prompt token count of ${latestMessage.tokenCount} exceeds max token count of ${this.maxContextTokens}.`,
-      );
+      const info = `${latestMessage.tokenCount} / ${this.maxContextTokens}`;
+      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
+      logger.warn(`Prompt token count exceeds max token count (${info}).`);
+      throw new Error(errorMessage);
     }
 
     if (usePrevSummary) {
@@ -410,19 +430,23 @@ class BaseClient {
       maxContextTokens: this.maxContextTokens,
     });
 
-    let tokenCountMap = orderedWithInstructions.reduce((map, message, index) => {
-      const { messageId } = message;
-      if (!messageId) {
-        return map;
-      }
+    /** @type {Record<string, number> | undefined} */
+    let tokenCountMap;
+    if (buildTokenMap) {
+      tokenCountMap = orderedWithInstructions.reduce((map, message, index) => {
+        const { messageId } = message;
+        if (!messageId) {
+          return map;
+        }
 
-      if (shouldSummarize && index === summaryIndex && !usePrevSummary) {
-        map.summaryMessage = { ...summaryMessage, messageId, tokenCount: summaryTokenCount };
-      }
+        if (shouldSummarize && index === summaryIndex && !usePrevSummary) {
+          map.summaryMessage = { ...summaryMessage, messageId, tokenCount: summaryTokenCount };
+        }
 
-      map[messageId] = orderedWithInstructions[index].tokenCount;
-      return map;
-    }, {});
+        map[messageId] = orderedWithInstructions[index].tokenCount;
+        return map;
+      }, {});
+    }
 
     const promptTokens = this.maxContextTokens - remainingContextTokens;
 
@@ -535,13 +559,19 @@ class BaseClient {
       isEdited,
       model: this.modelOptions.model,
       sender: this.sender,
-      text: addSpaceIfNeeded(generation) + completion,
       promptTokens,
       iconURL: this.options.iconURL,
       endpoint: this.options.endpoint,
       ...(this.metadata ?? {}),
     };
 
+    if (typeof completion === 'string') {
+      responseMessage.text = addSpaceIfNeeded(generation) + completion;
+    } else if (completion) {
+      responseMessage.text = '';
+      responseMessage.content = completion;
+    }
+
     if (
       tokenCountMap &&
       this.recordTokenUsage &&
@@ -861,8 +891,12 @@ class BaseClient {
 
           processValue(nestedValue);
         }
-      } else {
+      } else if (typeof value === 'string') {
         numTokens += this.getTokenCount(value);
+      } else if (typeof value === 'number') {
+        numTokens += this.getTokenCount(value.toString());
+      } else if (typeof value === 'boolean') {
+        numTokens += this.getTokenCount(value.toString());
       }
     };