VideoGen Advisor
Provides tools for generating cinematic videos, animating images into videos, and monitoring job status via Google Veo API.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@VideoGen Advisor Create a video where a presenter explains our Q4 results"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
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 selectionFull Control: Direct access to service-specific tools when needed
Related MCP server: video-mcp
Installation
pip install m2ai-mcp-videogen-advisorConfiguration
Set your API keys as environment variables:
export HEYGEN_API_KEY=your_heygen_api_key_here
export GEMINI_API_KEY=your_gemini_api_key_hereClaude 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 with intelligent routing to HeyGen or Veo |
| Check job status (works for both services) |
| Get download URL for completed video |
HeyGen-Specific (6 tools)
Tool | Description |
| List available avatars (stock + custom) |
| List available voices for avatar videos |
| Create video with specific avatar/voice |
| List available video templates |
| Get template variables and details |
| Generate video from template |
Veo-Specific (2 tools)
Tool | Description |
| Generate cinematic video with full Veo controls |
| 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
Generated by GRIMLOCK MCP Factory
Available Tools
11 toolscreate_avatar_videoB
Create a HeyGen avatar video with specific avatar and voice selection. Use for precise control over avatar.
Inputs: script (required), voice_id (required), avatar_id (required), test_mode, +2 more.
download_videoA
Get the download URL for a completed video. Call get_video_status first to ensure completion.
Inputs: job_id (required), service (required).
generate_creative_videoA
Generate cinematic video using Google Veo. Best for b-roll, visual storytelling, and scenes without a presenter.
Inputs: prompt (required), seed, model, duration, +3 more.
generate_from_templateB
Generate a HeyGen video from a template with custom variable values.
Inputs: variables (required), template_id (required), title, test_mode.
generate_videoA
Generate a video from a text prompt. Automatically routes to HeyGen (avatar/presenter) or Veo (creative/cinematic) based on content. Returns job_id for status polling.
Inputs: prompt (required), duration, aspect_ratio, service_hint.
generate_video_from_imageA
Animate a static image into video using Google Veo. The image becomes the first frame.
Inputs: prompt (required), image_url (required), model, duration, +1 more.
get_template_detailsB
Get HeyGen template details including all customizable variables.
Inputs: template_id (required).
get_video_statusA
Check the status of a video generation job. Works for both HeyGen and Veo. Poll until status is 'completed' or 'failed'.
Inputs: job_id (required), service (required).
list_avatarsA
List available HeyGen avatars including stock and custom Instant Avatars.
Inputs: filter_gender, include_instant.
list_templatesC
List available HeyGen templates with customizable variables.
Inputs: no named parameters.
list_voicesA
List available HeyGen voices for avatar videos including stock and cloned voices.
Inputs: filter_gender, filter_language.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 11 tools
Most tools have distinct purposes, but `generate_video` overlaps with `create_avatar_video` and `generate_creative_video` as it auto-routes to either engine, potentially confusing an agent.
Naming follows a verb_noun pattern with underscores, but there is a mix of 'create' and 'generate' for similar actions (e.g., `create_avatar_video` vs `generate_creative_video`), causing minor inconsistency.
With 11 tools, the server covers the core workflows of video generation, template management, and status tracking without being overly large or sparse.
The tool set covers creation, status polling, and downloading, but lacks a tool to list existing videos or manage them beyond status, leaving a minor gap in lifecycle management.
Related MCP Connectors
MCP server for Google Veo AI video generation
AI image, video, voice and music generation over MCP, routed to Veo 3.1, Seedance 2.5 and more.
MCP server for Wan AI video generation
AI video, image and audio generation with 140+ AI models from any MCP agent
Related MCP Servers
- AlicenseBqualityCmaintenanceMCP server that exposes Google's Veo2 video generation capabilities, allowing clients to generate videos from text prompts or images.735MIT
- AlicenseNot gradedqualityDmaintenanceAn async video generation MCP server with multi-provider support. Currently in skeleton phase with stub implementations, it will eventually enable video generation through providers like Veo 3.1, Grok Imagine Video, and Sora 2 Pro.MIT
- AlicenseNot gradedqualityBmaintenanceMCP server for 4K video generation using Google VEO 3.1 — text-to-video, image-to-video, video extension, and frame interpolation.3MIT
- AlicenseNot gradedqualityBmaintenanceMCP server for generating images and videos using Google Gemini and VEO models, with support for multiple AI models and credential modes.289 PyPI1Apache 2.0