MChat is a lightweight, embeddable, multi-tenant vertical RAG platform. It combines a streaming Bot engine, RAG knowledge base, Skill plugin system, and an embeddable chat Widget — with support for 10+ LLM providers and multi-channel connectivity (Web Widget, WebSocket, REST API, WeChat Official Account, and more).
The platform ships with AI customer service as a built-in base channel, and supports custom vertical channels — pre-configured domain RAG packages (e.g. patent search, medical consultation, legal Q&A) that bundle domain knowledge bases, specialized skill packs, and tuned retrieval strategies. Embed with a single <script> tag.
- English main site
- Chinese main site
- Full screenshot tour
- Product roadmap (knowledge base, Widget, channels, API, ops, RBAC)
Click any screenshot to open the full image.
- Bot engine — Streaming LLM inference + tool calling; OpenAI, Anthropic, Google, DeepSeek, Ollama, Groq, and more
- Skill plugins — Hot-reload
SKILL.mdpackages from disk/zip/URL, including OpenClaw-compatible formats. Premium skill packs available as vertical channel add-ons - RAG knowledge base — Multi-strategy chunking, multi-provider embeddings (OpenAI / local / Ollama), hybrid retrieval (vector + BM25 + RRF), multi-provider rerank, query rewriting, parent-child context
- Embeddable Widget — One
<script>tag for branded vertical RAG chat on any website - Multi-tenant — Independent channel configurations with isolated AI config, skills, and knowledge bases
- Vertical channels — Pre-configured domain RAG packages: AI model, system prompt, knowledge base, rerank strategy, skill packs, widget theme — one-click creation
- Multi-channel — Web Widget, REST, WebSocket, WeChat Official Account (DingTalk/WhatsApp/Telegram planned)
- Speech input — Voice-to-text via OpenAI Whisper (optional local models)
- Security — JWT authentication, API key management, RBAC
- Docker —
docker compose up -dfor full stack deployment
git clone https://github.com/windinwing/mchat.git
cd mchat
docker compose -f ops/docker/docker-compose.lite.yml up -d
# Admin UI: http://localhost:5173
# API docs: http://localhost:3001/docs
# Landing: http://localhost:5173/Default admin credentials (created on first startup): admin / admin123
Change the password under Admin → Users after sign-in. Override via ADMIN_USERNAME / ADMIN_PASSWORD in .env. Set SHOW_BOOTSTRAP_CREDENTIALS=false to hide the hint on the login page in production.
<script
src="http://localhost:5173/widget-loader.js"
data-mchat-url="http://localhost:3001"
data-agent-id="YOUR_AGENT_ID"
data-primary-color="#3b82f6"
data-welcome-message="Hello! How can I help you?"
data-bot-name="Assistant"
></script>make install # install dependencies
make dev # start backend + frontend
# Backend: http://localhost:3001 (/docs for Swagger)
# Frontend: http://localhost:5173make test # run tests
make lint # lint
mchat run # start via CLImchat/
├── src/
│ ├── backend/ # FastAPI (Python 3.12+)
│ │ └── app/
│ │ ├── api/ # REST routes
│ │ ├── bot/ # Bot engine + LLM providers
│ │ ├── knowledge/# RAG + Milvus
│ │ ├── skill/ # Skill system
│ │ └── channels/ # WeChat & channel adapters
│ └── frontend/ # React + Vite + Tailwind
│ └── src/
│ ├── i18n/ # zh / en (react-i18next)
│ └── pages/ # Landing + admin console
├── skills/ # Skill packages
├── channel_templates/ # Vertical channel templates (patent, medical, etc.)
├── docs/ # Architecture, API, deployment
├── ops/docker/ # Docker Compose
└── Makefile
| Provider | Default API base | Notes |
|---|---|---|
openai |
https://api.openai.com/v1 | GPT-4o, o1, etc. |
anthropic |
https://api.anthropic.com | Claude |
google |
https://generativelanguage.googleapis.com | Gemini |
deepseek |
https://api.deepseek.com | OpenAI-compatible |
ollama |
http://localhost:11434/v1 | Local models |
groq |
https://api.groq.com/openai/v1 | Fast inference |
zhipu / moonshot / siliconflow / together |
(configure api_base) |
Regional / hosted APIs |
openai-compatible |
(configure api_base) |
Any OpenAI-compatible endpoint |
| Group | Path | Description |
|---|---|---|
| Chat | /api/chat/* |
Conversations and messages |
| Agents | /api/agents/* |
AI configuration |
| Knowledge | /api/knowledge/* |
Documents and retrieval |
| Widget | /api/widget/* |
Embedded chat API |
| Skills | /api/skills/* |
Skill management |
| Channels | /api/channels/* |
WeChat and other channels |
| Channel Templates | /api/channels/templates/* |
One-click vertical channel creation |
| Speech | /api/speech/* |
Voice transcription |
| Auth | /api/auth/* |
Login / JWT |
| WebSocket | /ws |
Real-time streaming |
| Health | /api/health |
Service status |
See docs/api.md or /docs (Swagger) after startup.
The admin console and landing page support English and 简体中文. Language preference is stored in localStorage (mchat_lang). Switch language from the header or sidebar.
mchat init
mchat run
mchat config show
mchat skill list
mchat skill create <name>
mchat skill install <url-or-name>
mchat db init
mchat db seed- Supports standard frontmatter
SKILL.mdskill packages - Supports OpenClaw-style
SKILL.mdlocale blocks - Admin can install skills by zip upload or URL (
/api/skills/install-url) - CLI supports direct URL or ClawHub name install, for example:
mchat skill install patent-search - Premium skill packs: Vertical channels can bundle specialized skills as value-added services
| File | Services | Use case |
|---|---|---|
docker-compose.lite.yml |
MySQL + Backend + Frontend | Dev / lightweight |
docker-compose.yml |
+ Milvus, etcd, MinIO, Redis | Full RAG |
docker-compose.prod.yml |
+ Nginx HTTPS | Production |
docker-compose.dev.yml |
Hot reload | Local dev |
Backend: FastAPI, SQLAlchemy 2.0, Milvus, OpenAI / Anthropic SDKs, JWT, Loguru
Frontend: React 19, TypeScript, Vite, Tailwind CSS 4, Zustand, react-i18next
Contributions are welcome. See CONTRIBUTING.md if present.