Open Avatar Chat

中文 | English

A modular interactive digital human conversation implementation.

🤗 Demo&nbsp&nbsp|&nbsp&nbsp Demo&nbsp&nbsp|&nbsp&nbsp💬 WeChat&nbsp&nbsp|&nbsp&nbsp📖 Docs

💡 Core Highlights

• Multimodal Interaction: Supports text, audio, video and other interaction methods for natural human-machine dialogue
• Modular Architecture: Highly modular design with flexible ASR, LLM, TTS, and Avatar component replacement
• Diverse Avatar Options: Supports LiteAvatar, LAM, MuseTalk, FlashHead and other digital human technologies
• Low Latency: Optimized through VAD detection, audio buffering, and frame rate control with ~2.2s average response time

📢 News

• [2026.04] ⭐️⭐️⭐️ Version 0.6.0 Released:
• Architecture refactored with frontend/backend separation: OpenAvatarChat-WebUI
• All avatars now support manual interrupt and duplex interrupt modes
• Optimized installation, deployment, and model download workflow
• Integrated SoulX-FlashHead diffusion-based real-time streaming talking head
• [2025.08.19] ⭐️⭐️⭐️ Version 0.5.1 Released:
• LiteAvatar multi-session support
• Added Qwen-Omni multimodal model support

📋 Full Release Notes

Demo

Try it Online

We have deployed demo services on ModelScope and HuggingFace. Feel free to try it out.

Demo Video

Component Dependencies

Type	Open Source Project	GitHub Link	Model Link
RTC	HumanAIGC-Engineering/gradio-webrtc
WebUI	HumanAIGC-Engineering/OpenAvatarChat-WebUI
VAD	snakers4/silero-vad
Avatar	HumanAIGC/lite-avatar
TTS	FunAudioLLM/CosyVoice
Avatar	aigc3d/LAM_Audio2Expression		🤗
	facebook/wav2vec2-base-960h		🤗
Avatar	TMElyralab/MuseTalk
Avatar	Soul-AILab/SoulX-FlashHead		🤗

🚀 Quick Start

# Clone the project
git clone https://github.com/HumanAIGC-Engineering/OpenAvatarChat.git
cd OpenAvatarChat
git submodule update --init --recursive --depth 1

# Install dependencies (LiteAvatar + Bailian API example)
uv run install.py --config config/chat_with_openai_compatible_bailian_cosyvoice.yaml

# Download models
uv run scripts/download_models.py --handler liteavatar

# Start
uv run src/demo.py --config config/chat_with_openai_compatible_bailian_cosyvoice.yaml

📖 See Getting Started for detailed instructions.

Preset Modes

📖 View all preset modes

🧪 Beta Features

Chat Agent Mode (OpenClaw Integration)

[!WARNING] This feature is currently in Beta. APIs and configuration formats may change at any time.

Chat Agent mode replaces the traditional LLM Handler with a multi-turn tool-calling Agent, providing:

• Tool Calling: Invoke tools across multiple turns (get time, system info, etc.)
• Persona & Long-term Memory: Persistent persona through OpenClaw's Agent Profile
• Context Compression: Automatically compresses long conversation history
• Background Task Collaboration: Execute complex tasks via OpenClaw in the background
• Visual Perception: Camera input processing via PerceptionAgent

📖 Full Chat Agent documentation

Community

• WeChat Group

• 🚨 FAQ

Community Thanks

• One-click installation package by "Shi Zi Yu" on Bilibili
• Quark one-click packages by "W&H": Windows (code: a79V) / Linux (code: E8Kq)
• Source code zip by "W&H": Quark (code: 9iNy) / Baidu (code: xrxr)

Star History

Citation

If you found OpenAvatarChat helpful in your research/project, we would appreciate a Star⭐ and citation✏️

@software{avatarchat2025,
  author = {Gang Cheng, Tao Chen, Feng Wang, Binchao Huang, Hui Xu, Guanqiao He, Yi Lu, Shengyin Tan},
  title = {OpenAvatarChat},
  year = {2025},
  publisher = {GitHub},
  url = {https://github.com/HumanAIGC-Engineering/OpenAvatarChat}
}