Genrelay Mcp

Lets an assistant generate videos and images through the GenRelay API using models such as Veo 3.1, Grok Imagine and GPT Image 2.

What you can do

Exposes four tools: generate_video for text-to-video and image-to-video, generate_image for image generation and editing (returned inline), check_job to retrieve a job by task id, and list_models to see available models. Video jobs often exceed the tool-call time budget, so generate_video returns a task id and progress instead of the clip, which is then collected with check_job. Several video and image models are selectable, with quality tiers for images and Veo video.

When to use it

Use it when you want an MCP client such as Claude Desktop or Cursor to create video clips or images from prompts without leaving the chat, including image-to-video, first/last-frame video, and image editing. It suits occasional creative generation rather than bulk or automated pipelines.

Requirements

Runs locally over stdio, launched with npx -y genrelay-mcp, so Node.js and npx are needed. A GenRelay account and API key are required, supplied as the GENRELAY_API_KEY environment variable; GENRELAY_BASE_URL is optional and defaults to the GenRelay API endpoint. Network access to the GenRelay service is required, and the client must be restarted after configuration changes.