generate_text_to_videoLABSCO SUMMARY
Two video models behind one conversation.
RunwayML and Luma AI have different strengths and separate dashboards. Putting both behind one MCP server means a prompt can be tried on either without leaving the chat, and Luma's follow-up steps — upscale, add audio — happen on the generation you just made.
What it is
An MCP server that fronts the RunwayML and Luma AI APIs for video and image generation, so a generation request, its settings and its result stay inside the assistant you are already using.
What you get
- Text-to-video on either provider, with model, aspect ratio, resolution and watermark chosen per request
- Image-to-video, taking a still you supply plus a prompt describing the motion
- Image generation from text through Luma AI
- Luma generation management — list, fetch and delete — plus upscaling and audio added to an existing generation
- Prompt enhancement through an OpenRouter model before the generation request is sent
Requirements
Node.js v18 LTS or later, then clone, npm install and npm run build. Keys for RunwayML, Luma AI and OpenRouter — the last of those only for the prompt-enhancement step — go in a .env or the client config. Generation is billed by RunwayML and Luma on their own terms.
Setup effort
Build from source — git clone https://github.com/wheattoast11/mcp-video-gen.git
USE CASES
01
Getting a clip out of a written idea
Describe the shot and pick the provider, aspect ratio and resolution in the same request rather than filling in a web form.
02
Putting motion into a still you already have
Supply the image and a prompt for how it should move, keeping the frame you approved instead of regenerating from scratch.
03
Comparing the same prompt on two models
Run one description through RunwayML and Luma AI and judge the outputs side by side before committing budget to a batch.
04
Sharpening a prompt before you spend a generation
Pass the description through an OpenRouter model first, so the request that costs money is the improved one.
05
Finishing a Luma clip you want to keep
Upscale it and add audio on the existing generation, and clear out the attempts you do not want.
10 tools
Listed in project docs. Your client may show a different list.
generate_image_to_videoenhance_promptluma_generate_imageluma_list_generationsluma_get_generationluma_delete_generationluma_get_camera_motionsluma_add_audioluma_upscaleBEFORE YOU INSTALL
What to know first
Tools you get
10 tools named in the project's documentation.
Requires
- Node.js
- RUNWAYML_API_SECRET in your environment
- LUMAAI_API_KEY in your environment
- OPENROUTER_API_KEY in your environment
Where it runs
Both documented clients run the server on your machine.
INSTALLATION
Connect your client
You clone and build this one yourself.
Server setup
TERMINAL
git clone https://github.com/wheattoast11/mcp-video-gen.git
cd mcp-video-gen
npm install
npm run buildMCP Clients
Claude Desktop, Cline
JSON
{
"mcpServers": {
"runway-luma-server": {
"command": "node",
"args": [
"<YOUR_PATH_TO_MCP_VIDEO_GEN>/build/server-index.js"
],
"env": {
"RUNWAYML_API_SECRET": "<YOUR_RUNWAYML_API_SECRET>",
"LUMAAI_API_KEY": "<YOUR_LUMAAI_API_KEY>",
"OPENROUTER_API_KEY": "<YOUR_OPENROUTER_API_KEY>"
},
"disabled": false,
"autoApprove": []
}
}
}
