pingLABSCO SUMMARY
The whole Dream Machine loop, including the waiting.
Generation is asynchronous, so get_generation is the call you will make most: it reports queued, dreaming, completed or failed, and gives the reason when it fails. Keyframes are the interesting argument, because pointing frame0 or frame1 at an earlier generation is how you extend or bridge clips instead of starting over.
What it is
A server for Luma AI's Dream Machine API v1, covering video and image generation, generation management and credits.
What you get
- Text-to-video generation with model, resolution up to 4k, a choice of clip length, aspect ratio and looping (create_generation)
- Keyframes on that same call, frame0 and frame1 taken from an image URL or an earlier generation, which is how extension and interpolation are done
- Image generation with reference images, style references and negative prompts (generate_image)
- A generation's state, queued, dreaming, completed or failed, with the failure reason or the finished video URL (get_generation)
- Your generation history listed with limit and offset, and generations deleted (list_generations, delete_generation)
- A finished video upscaled to a higher resolution (upscale_generation), and audio added to it (add_audio)
Requirements
A Luma AI API key for the Dream Machine API.
Setup effort
Build from source
USE CASES
01
Turning a written idea into a video clip
Create the generation with the model, resolution and aspect ratio you want.
02
Finding out whether a generation finished or failed
Check its state and read either the video URL or the failure reason.
03
Extending or bridging clips you already made
Set a keyframe to an earlier generation so the new clip continues from it.
04
Making a still image from a reference you supply
Generate with an image reference, a style reference or a negative prompt.
05
Finishing a clip for delivery
Upscale it to a higher resolution and add the audio track.
06
Clearing out generations you no longer want
List your history and delete what is not worth keeping.
10 tools
Listed in project docs. Your client may show a different list.
create_generationget_generationlist_generationsdelete_generationupscale_generationadd_audiogenerate_imageget_creditsget_camera_motionsBEFORE YOU INSTALL
What to know first
Tools you get
10 tools named in the project's documentation.
Requires
LUMA_API_KEY in your environment
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone this one yourself.
Server setup
TERMINAL
git clone https://github.com/bobtista/luma-ai-mcp-server.gitMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"luma": {
"command": "uv",
"args": [
"run",
"--project",
"<YOUR_PATH_TO_LUMA_AI_MCP_SERVER>",
"-m",
"luma_ai_mcp_server"
],
"env": {
"LUMA_API_KEY": "<YOUR_LUMA_API_KEY>"
}
}
}
}
