ChatMCP

Stale
GitHub Dart Apache-2.0

Description

AI chat client implementing the Model Context Protocol (MCP) with multi-model support and cross-platform desktop experience

Key Features

  • Cross-platform AI chat client for macOS, Windows, Linux, iOS, Android, and Web with Flutter
  • Full MCP protocol support with stdio and SSE transport modes for tool calling
  • Multi-model provider support including OpenAI-compatible APIs, DeepSeek, Ollama, and more
  • Artifact display, thinking mode, image generation, LaTeX, HTML preview, and Mermaid diagrams
  • LAN data sync between ChatMCP instances for seamless cross-device experience
  • Built-in MCP server management UI for installing and configuring tool servers

Use Cases

💡 Unified AI chat experience across desktop and mobile with MCP tool integration
💡 Connecting to custom LLM providers while using MCP-powered tools and workflows
💡 Sharing chat history and settings across multiple devices on the same network
💡 Developing and testing MCP servers with a visual debugging interface
💡 Running AI assistants with artifact rendering capabilities for rich content output

Categories

Quick Start

1. Install uv (brew install uv) or npx (brew install node)
2. Download the release for your platform from https://github.com/daodao97/chatmcp/releases
3. Configure your LLM API key and endpoint in the Settings page
4. Install MCP servers from the MCP Server page
5. Start chatting with MCP-powered tools

Related Projects