# ChatJS Stop rebuilding the same AI chat infrastructure. ChatJS gives you a production-ready foundation with authentication, 120+ models, streaming, and tools so you can focus on what makes your app unique. [**Website**](https://chatjs.dev) · [**Live Demo**](https://demo.chatjs.dev) · [**Documentation**](https://chatjs.dev/docs) ![DemosOnly](https://github.com/user-attachments/assets/f12e89dd-c10c-4e06-9b1a-a9fbd809d234)

## CLI Create a new ChatJS app: ```bash npx @chat-js/cli@latest create my-app ``` The CLI walks you through gateway, features, and auth choices, generates `chat.config.ts`, and lists the env vars required by your selections. ## Features - **120+ Models**: Claude, GPT, Gemini, Grok via one API - **Auth**: GitHub, Google, anonymous. Ready to go. - **Attachments**: Images, PDFs, docs. Drag and drop. - **Resumable Streams**: Continue generation after page refresh - **Branching**: Fork conversations, explore alternatives - **Sharing**: Share conversations with public links - **Web Search**: Real-time web search integration - **Image Generation**: AI-powered image creation - **Code Execution**: Run code snippets in sandbox - **MCP**: Model Context Protocol support - **Desktop App**: Package as a native macOS, Windows, or Linux app with Electron ## Stack - [Next.js](https://nextjs.org) - App Router, React Server Components - [TypeScript](https://www.typescriptlang.org) - Full type safety - [AI SDK](https://ai-sdk.dev/) - The AI Toolkit for TypeScript - [AI Gateway](https://vercel.com/ai-gateway) - Unified access to 120+ AI models - [Better Auth](https://www.better-auth.com) - Authentication & authorization - [Drizzle ORM](https://orm.drizzle.team) - Type-safe database queries - [PostgreSQL](https://www.postgresql.org) - Primary database - [Redis](https://redis.io) - Caching & resumable streams - [Vercel Blob](https://vercel.com/storage/blob) - Blob storage - [Shadcn/UI](https://ui.shadcn.com) - Beautiful, accessible components - [Tailwind CSS](https://tailwindcss.com) - Styling - [tRPC](https://trpc.io) - End-to-end type-safe APIs - [Zod](https://zod.dev) - Schema validation - [Zustand](https://zustand.docs.pmnd.rs/) - State management - [Motion](https://motion.dev) - Animations - [t3-env](https://env.t3.gg) - Environment variables - [Pino](https://getpino.io) - Structured Logging - [Langfuse](https://langfuse.com) - LLM observability & analytics - [Vercel Analytics](https://vercel.com/analytics) - Web analytics - [Biome](https://biomejs.dev) - Code linting and formatting - [Ultracite](https://ultracite.ai) - Biome preset for humans and AI - [Streamdown](https://streamdown.ai/) - Markdown for AI streaming - [AI Elements](https://elements.ai-sdk.dev/overview) - AI-native Components - [AI SDK Tools](https://ai-sdk-tools.dev/) - Developer tools for AI SDK ## Monorepo Layout - `apps/site`: Landing page ([chatjs.dev](https://chatjs.dev)) - `apps/chat`: Next.js chat app ([demo.chatjs.dev](https://demo.chatjs.dev)) - `apps/docs`: Blume docs ([chatjs.dev/docs](https://chatjs.dev/docs)) - `packages/cli`: interactive scaffold CLI ## Development - `bun dev`: run chat app - `bun dev:info`: print this worktree's assigned app URLs - `bun dev:docs`: run docs - `bun lint`: run workspace lint - `bun test:types`: run chat app typecheck Set `CHATJS_DEV_SLOT` in `.env.worktree.local` to reserve a stable range of ten ports per worktree. Within each range, chat uses offset `0`, Electron uses `1`, and the site uses `2`, as configured in `.worktree-env.json`. The local file is ignored by Git and kept separate from Vercel-managed `.env.local`. Run `bun dev:info` instead of assuming a port. ## Releases Releases are driven by Changesets for the whole repository. - Add a changeset for each releasable package you change. - Merge the generated version PR from the Changesets workflow. - Public packages such as `@chat-js/cli` publish to npm. - Desktop installers for `@chat-js/electron` publish to GitHub Releases. ## License Apache-2.0