VideoGen Advisor
A unified MCP server for video generation that intelligently routes requests to HeyGen (for avatar/presenter videos) or Google Veo (for creative/cinematic content).
README
VideoGen Advisor
A unified MCP server for video generation that intelligently routes requests to either HeyGen (for avatar/presenter videos) or Google Veo (for creative/cinematic content).
Features
- Intelligent Routing: Automatically selects the best service based on prompt content
- HeyGen Integration: Avatar-based videos, custom voices, templates
- Google Veo Integration: Cinematic content, image-to-video, creative visuals
- Unified Interface: Single
generate_videotool with automatic service selection - Full Control: Direct access to service-specific tools when needed
Installation
pip install m2ai-mcp-videogen-advisor
Configuration
Set your API keys as environment variables:
export HEYGEN_API_KEY=your_heygen_api_key_here
export GEMINI_API_KEY=your_gemini_api_key_here
Claude Desktop Configuration
Add to your claude_desktop_config.json:
{
"mcpServers": {
"videogen": {
"command": "videogen-advisor",
"env": {
"HEYGEN_API_KEY": "your_heygen_key",
"GEMINI_API_KEY": "your_gemini_key"
}
}
}
}
Or run directly with Python:
{
"mcpServers": {
"videogen": {
"command": "python",
"args": ["-m", "videogen_mcp.server"],
"env": {
"HEYGEN_API_KEY": "your_heygen_key",
"GEMINI_API_KEY": "your_gemini_key"
}
}
}
}
Available Tools
Unified Interface (3 tools)
| Tool | Description |
|---|---|
generate_video |
Generate video with intelligent routing to HeyGen or Veo |
get_video_status |
Check job status (works for both services) |
download_video |
Get download URL for completed video |
HeyGen-Specific (6 tools)
| Tool | Description |
|---|---|
list_avatars |
List available avatars (stock + custom) |
list_voices |
List available voices for avatar videos |
create_avatar_video |
Create video with specific avatar/voice |
list_templates |
List available video templates |
get_template_details |
Get template variables and details |
generate_from_template |
Generate video from template |
Veo-Specific (2 tools)
| Tool | Description |
|---|---|
generate_creative_video |
Generate cinematic video with full Veo controls |
generate_video_from_image |
Animate a static image into video |
Usage Examples
Automatic Routing
"Create a video where someone presents our Q4 results"
→ Routes to HeyGen (presenter/speaking content)
"Cinematic aerial shot of mountains at sunset"
→ Routes to Veo (creative/visual content)
Explicit Service Selection
# Force HeyGen
generate_video(
prompt="Create any video",
service_hint="heygen"
)
# Force Veo
generate_video(
prompt="A tutorial video",
service_hint="veo"
)
HeyGen Avatar Video
# 1. List avatars and voices
avatars = list_avatars(filter_gender="female")
voices = list_voices(filter_language="en")
# 2. Create video
job = create_avatar_video(
script="Hello! Welcome to our company.",
avatar_id="Angela-inblackskirt-20220820",
voice_id="1bd001e7e50f421d891986aad5158bc8",
background={"type": "color", "value": "#0066CC"},
aspect_ratio="16:9"
)
# 3. Poll for completion
status = get_video_status(job_id=job["job_id"], service="heygen")
# 4. Download when complete
result = download_video(job_id=job["job_id"], service="heygen")
Veo Creative Video
# Generate cinematic content
job = generate_creative_video(
prompt="A slow dolly shot through a neon-lit Tokyo alley at night, rain reflections, cinematic color grading",
negative_prompt="blurry, low quality, text overlays",
aspect_ratio="16:9",
duration=8,
model="veo-3.1-generate-preview"
)
# Poll and download
status = get_video_status(job_id=job["job_id"], service="veo")
result = download_video(job_id=job["job_id"], service="veo")
Image to Video
# Animate a still image
job = generate_video_from_image(
image_url="https://example.com/product-photo.jpg",
prompt="Slow zoom in with subtle product rotation",
duration=6
)
Routing Logic
The generate_video tool uses keyword detection to route requests:
HeyGen triggers (avatar/presenter content):
- speak, present, explain, tutorial, avatar, host, narrator
- talking, announce, introduce, walk through, demo, spokesperson
Veo triggers (creative/cinematic content):
- cinematic, scene, visual, b-roll, animation, shot, footage
- landscape, abstract, artistic, dramatic, aerial, timelapse
Ambiguous prompts default to Veo unless avatar/presenter language is detected.
Video Wait Times
- HeyGen: 1-5 minutes (varies by content length)
- Veo: 2-10 minutes (varies by complexity)
Poll get_video_status until status is completed or failed.
Error Handling
All tools return structured error objects:
{
"error": true,
"code": "QUOTA_EXCEEDED",
"message": "API credits exhausted. Check your account."
}
Common error codes:
AUTH_FAILED- Invalid API keyQUOTA_EXCEEDED- API credits exhaustedINVALID_AVATAR/INVALID_VOICE- ID not foundCONTENT_BLOCKED- Safety filter triggeredVIDEO_NOT_READY- Generation still in progressJOB_NOT_FOUND- Invalid job ID
Development
# Run tests
pytest
# Run tests with coverage
pytest --cov=videogen_mcp
# Type checking
mypy src/
# Linting
ruff check src/ tests/
License
MIT License
Credits
- HeyGen API
- Google Veo API
- Generated by GRIMLOCK MCP Factory
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.