Run AI chats and tasks in familiar messengers
You can operate a bot that lets you chat with AI and handle tasks in familiar apps like QQ or Telegram. AstrBot is an open-source platform that connects various messengers and AI models, enabling you to build everything from a personal AI companion to an enterprise knowledge base. It offers the experience of installing over 1,000 plugins at once to add desired features directly to your chat window.
Connecting messengers and AI models
It connects broadly with major LLM services such as OpenAI, Anthropic, Google Gemini, and DeepSeek, as well as self-hosted models like Ollama. Integration with agent platforms like Dify, Alibaba Cloud Bailian, and Coze is also possible, allowing you to bring existing AI workflows directly into a chat interface. It supports voice recognition and synthesis services, enabling voice-based interaction in addition to text.
Safe execution environment and scalability
To reduce risks when AI executes code or calls shell commands, it provides an agent sandbox. This sandbox performs tasks in an isolated environment, allowing session-level resource reuse without affecting the entire system. You can install over 1,000 plugins with one click, making it easy to extend the chatbot's functionality. It includes a web-based management interface and chat UI, so it can be used easily on a desktop rather than just in a server environment.
Installation methods and requirements
The fastest way to start is using the uv tool. For macOS users, security checks may add 10 to 20 seconds to the first run. For more stable operation, deployment via Docker or Docker Compose is recommended. If you primarily use ChatUI on a desktop, use AstrBot App; if you want quick deployment and multi-instance management, use AstrBot Launcher.
Pre-use checks
AstrBot is based on a Python 3.12 environment and requires a uv or Docker environment. It is advisable to check in advance for official support of specific messenger platforms and model service compatibility. For example, WhatsApp is currently in a 'Coming Soon' state. If you wish to contribute to the community, it is important to follow code formatting and linting rules using ruff. This project targets individuals, developers, and teams alike, aiming to build AI applications that can be applied directly to production environments.
