AstrBot
An easy-to-deploy, extensible open-source AI chatbot
Features
System Requirements
Minimum 8GB RAM. Low storage requirement (5GB reserved). Windows 10/11 64-bit: No specific GPU requirements. macOS 11+: Supports both Intel and M-series chips.
Introduction
Note: The default username is "astrbot" and the password is "astrbot". Please change the username and password after the first login.
AstrBot is a highly extensible, cross-platform, multi-model intelligent chatbot framework maintained by the AstrBotDevs open-source development team. Designed for general users, developers, enterprise office scenarios, and community operations, it is built on a modern asynchronous architecture that allows anyone to quickly deploy a powerful and customizable AI robot without advanced programming skills.
1. Core Features for Regular Users
Multi-platform messaging support AstrBot can connect to QQ, Telegram, WeChat Work, Lark, DingTalk, Discord, and other mainstream social and office platforms. A single bot can manage conversations across all platforms simultaneously.
Access to almost all major large language models It supports OpenAI, DeepSeek, Gemini, Tongyi Qianwen, ERNIE, Moonshot AI, SiliconFlow, and many others. It also works with Ollama, OneAPI, Dify, and other proxy or local deployment solutions. Users can switch models freely without modifying code.
Complete dialogue system & customizable personality You can customize the bot’s tone, style, role, and behavior. It supports long-term conversation memory to maintain natural, continuous multi-turn dialogues.
Multi-modal capabilities (text, image, voice)
- Image recognition and vision-language dialogue
- Built-in speech-to-text (Whisper, SenseVoice)
- Text-to-speech synthesis (OpenAI TTS, Edge TTS, etc.) Users can chat using voice messages, and the bot can reply with voice.
Powerful Agent capabilities With built-in tool calling, the bot can autonomously:
- Search the web for real-time information
- Run code for calculations and data analysis
- Use external plugins
- Query local knowledge bases for professional answers It acts as a truly thinking, actionable intelligent assistant.
Built-in security and management tools Includes content moderation, rate limits, blacklists/whitelists, permission control, and group management, making it safe for communities, schools, and businesses.
Extremely easy deployment Offers Windows one-click package, Docker images, 1Panel & BT Panel deployment. Regular users can finish setup in minutes without server or command-line knowledge.
2. Product Features & Advantages
- Highly modular & loosely coupled: Platforms, models, plugins, and messages are independent for easy extension.
- Rich plugin ecosystem: Supports third-party plugins for check-in, entertainment, queries, AI painting, scheduled tasks, etc.
- Native knowledge base: Upload documents directly to build a custom knowledge base without external systems.
- Web-based visual dashboard: Configure models, platforms, plugins, permissions, and personas via browser.
- High-performance async architecture: Based on Python async IO for high concurrency and low latency.
- 100% open-source & free: No hidden features or paywalls, with active community updates.
3. Typical Usage Scenarios
- Personal AI assistant for daily questions
- Community management bots (QQ groups, Telegram groups)
- Enterprise internal customer service & knowledge base Q&A
- Office automation (Lark, DingTalk, WeChat Work)
- Rapid AI prototyping for developers
- Educational and organizational Q&A systems