AI Video Generator

You have an MCP server that uses AI video generation models to turn text prompts or images into video clips. It lets you control video parameters, generate with starting or ending images, check progress, and switch between supported AI models for different visual results.

How to use

Use this MCP server with a compatible MCP client to generate videos from prompts or starting/ending images. You can select the AI model, adjust aspect ratio, resolution, and duration, and decide if the video should loop. You can also check the status of a video generation request by its ID.

How to install

Prerequisites: Node.js and npm need to be installed on your system.

npm install

Create a configuration file and set your FAL.AI API key as an environment variable.

FAL_KEY=your_fal_key_here

Build the server after installing dependencies.

npm run build

Run the server directly.

npm start

Additional configuration and usage notes

Claude Desktop integration is supported. To configure, add an MCP entry that points to the local build of the server and provides the API key in the environment.

{
  "mcpServers": {
    "video-generator": {
      "command": "node",
      "args": ["your_install_path/fal-mcp-server/build/index.js"],
      "env": {
        "FAL_KEY": "your_fal_key_here"
      }
    }
  }
}

Parameters and available tools

The server exposes tools to generate videos and to check their status.

generate-video uses AI models to create a video from text and/or an initial image and can apply optional end images, aspect ratio, resolution, duration, looping, and model choice.

check-video-status retrieves the status of a video generation request by its ID.

Claude-inspired usage example helps you see how prompts are turned into parameterized tool calls.

Model comparison and limits

Luma Ray2 Flash produces smooth motion and realistic physics-inspired results with natural movement.

Kling v1.6 Pro delivers detailed textures and stylized effects, offering more pronounced artistic results.

Video generation may take longer for high resolutions or longer durations, and you need a valid FAL.AI API key with sufficient credits. Higher resolutions and longer videos consume more credits.

Troubleshooting

API key errors usually mean the FAL_KEY environment variable is not set correctly. You can also set it directly in the Claude Desktop configuration.

If video generation fails, check for issues like an invalid API key, insufficient credits, inappropriate prompts or images, or temporary server problems. If persistent, retry after a pause or adjust the prompt.

License

MIT license.

Available tools

generate-video

Generates a video from a text prompt and/or a starting image, with options for aspect ratio, resolution, duration, looping, and model selection.

check-video-status

Checks the status of a video generation request using a request_id and the model used.