Skip to main content

HeyGen avatar plugin for Vision Agents

Project description

HeyGen Avatar Plugin for Vision Agents

Add realistic avatar video to your AI agents using HeyGen's streaming avatar API.

Features

  • 🎭 Realistic Avatars: Use HeyGen's high-quality avatars with natural movements
  • 🎤 Automatic Lip-Sync: Avatar automatically syncs with audio from any TTS provider
  • 🚀 WebRTC Streaming: Low-latency real-time video streaming via WebRTC
  • 🔌 Easy Integration: Works seamlessly with Vision Agents framework
  • 🎨 Customizable: Configure avatar, quality, resolution, and more

Installation

pip install vision-agents-plugins-heygen

Or with uv:

uv pip install vision-agents-plugins-heygen

Quick Start

import asyncio
from uuid import uuid4
from dotenv import load_dotenv

from vision_agents.core import User, Agent
from vision_agents.plugins import cartesia, deepgram, getstream, gemini, heygen
from vision_agents.plugins.heygen import VideoQuality

load_dotenv()

async def start_avatar_agent():
    agent = Agent(
        edge=getstream.Edge(),
        agent_user=User(name="AI Assistant with Avatar", id="agent"),
        instructions="You're a friendly AI assistant.",
        
        llm=gemini.LLM("gemini-2.0-flash"),
        tts=cartesia.TTS(),
        stt=deepgram.STT(),
        
        # Add HeyGen avatar
        processors=[
            heygen.AvatarPublisher(
                avatar_id="default",
                quality=VideoQuality.HIGH
            )
        ]
    )
    
    call = agent.edge.client.video.call("default", str(uuid4()))
    
    with await agent.join(call):
        await agent.simple_response("Hello! I'm your AI assistant with an avatar.")
        await agent.finish()

if __name__ == "__main__":
    asyncio.run(start_avatar_agent())

Configuration

Environment Variables

Set your HeyGen API key:

HEYGEN_API_KEY=your_heygen_api_key_here

AvatarPublisher Options

from vision_agents.plugins.heygen import VideoQuality

heygen.AvatarPublisher(
    avatar_id="default",           # HeyGen avatar ID
    quality=VideoQuality.HIGH,    # Video quality: VideoQuality.LOW, VideoQuality.MEDIUM, or VideoQuality.HIGH
    resolution=(1920, 1080),       # Output resolution (width, height)
    api_key=None,                  # Optional: override env var
)

Usage Examples

With Realtime LLM

from vision_agents.plugins import gemini, heygen, getstream

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="Realtime Avatar AI"),
    instructions="Be conversational and responsive.",
    
    llm=gemini.Realtime(fps=2),  # No separate TTS needed
    
    processors=[
        heygen.AvatarPublisher(avatar_id="professional_presenter")
    ]
)

call = agent.edge.client.video.call("default", str(uuid4()))

with await agent.join(call):
    await agent.finish()

With Multiple Processors

from vision_agents.plugins import ultralytics, heygen

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="Fitness Coach"),
    instructions="Analyze user poses and provide feedback.",
    
    llm=gemini.Realtime(fps=3),
    
    processors=[
        # Process incoming user video
        ultralytics.YOLOPoseProcessor(model_path="yolo11n-pose.pt"),
        # Publish avatar video
        heygen.AvatarPublisher(avatar_id="fitness_trainer")
    ]
)

How It Works

  1. Connection: Establishes WebRTC connection to HeyGen's streaming API
  2. Audio Input: Receives audio from your TTS provider or Realtime LLM
  3. Avatar Generation: HeyGen generates avatar video with lip-sync
  4. Video Streaming: Streams avatar video to call participants via GetStream Edge

Requirements

  • Python 3.10+
  • HeyGen API key (get one at heygen.com)
  • GetStream account for video calls
  • TTS provider (Cartesia, ElevenLabs, etc.) or Realtime LLM

Troubleshooting

Connection Issues

If you experience connection problems:

  1. Check your HeyGen API key is valid
  2. Ensure you have network access to HeyGen's servers
  3. Check firewall settings for WebRTC traffic

Video Quality

To optimize video quality:

  • Use quality=VideoQuality.HIGH for best results
  • Increase resolution if bandwidth allows
  • Ensure stable internet connection

API Reference

AvatarPublisher

Main class for publishing HeyGen avatar video.

Methods:

  • publish_video_track(): Returns video track for streaming
  • state(): Returns current state information
  • close(): Clean up resources

License

MIT

Links

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

vision_agents_plugins_heygen-0.2.3.tar.gz (12.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

vision_agents_plugins_heygen-0.2.3-py3-none-any.whl (33.1 kB view details)

Uploaded Python 3

File details

Details for the file vision_agents_plugins_heygen-0.2.3.tar.gz.

File metadata

File hashes

Hashes for vision_agents_plugins_heygen-0.2.3.tar.gz
Algorithm Hash digest
SHA256 e8924487db5fafb6bd04bbea3b9c92d03c964df9cc875b22ff780e9cbf98bf84
MD5 dba8fcca6ea7e307917f79b5040b8496
BLAKE2b-256 bdad3d6e370527f9a816f8aa9df55e1418499f1ffaa4b90adb6fcbfcbe84da0f

See more details on using hashes here.

File details

Details for the file vision_agents_plugins_heygen-0.2.3-py3-none-any.whl.

File metadata

File hashes

Hashes for vision_agents_plugins_heygen-0.2.3-py3-none-any.whl
Algorithm Hash digest
SHA256 3055b81bb8a89b66bf5b1ca6bcef4a3602a193e8a377ca21629a9f9a8066e77c
MD5 6d0f42f237af629b1ca6942b336fc712
BLAKE2b-256 9dfb4f01c246d1b09d950e88e92b211778c99d46b9756d554ef7a798cf08a944

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page