What is Fal MCP?
The Fal MCP server is a Model Context Protocol server that connects Fal to AI agents and assistants as a set of tools. With it connected, an agent can work with Fal (generative media inference — run image, video and audio models) directly — instead of you copying data in and out by hand.
It is a remote MCP server that uses OAuth, so you point your client at the server URL, authorize once, and the Fal tools become available to the model.
Connect the Fal MCP server
Connect Fal MCP to Claude Code
- 1
Add the server
Use `claude mcp add` with the Fal MCP server URL (or its `npx` command), following the official docs.
- 2
Authorize
Complete authentication (OAuth) so Claude Code can call the server.
- 3
Verify
Ask Claude to use a Fal tool to confirm the connection is live.
Connect Fal MCP to Cursor
- 1
Open MCP settings
In Cursor, go to Settings → MCP → Add new server.
- 2
Add Fal
Provide the Fal MCP server URL or command and complete authentication.
- 3
Test
Reference Fal from chat to confirm Cursor can reach the tools.
Sample use cases
An agent generates product mockup images from text descriptions and uploads them to a shared drive for the design team
Faster creative iteration with AI-generated visual concepts.
A content pipeline agent creates hero images for blog posts by running a prompt through Fal's image generation models
Every article ships with a unique, on-brand visual without waiting for a designer.
An agent upscales and enhances low-resolution product photos before they are published to an e-commerce catalog
Consistent, high-quality product imagery across all listings.
A scheduled agent runs inference on a fine-tuned model to classify incoming support tickets by category
Support tickets are routed to the right team instantly, reducing first-response time.
An agent converts rough wireframe sketches into polished UI mockups using an image-to-image model on Fal
Designers get a starting point for high-fidelity screens in seconds.
A marketing agent generates ad creative variations by prompting Fal with different copy and style directions
More creative options for A/B testing without additional design resources.
An agent transcribes and summarizes audio recordings using Fal's speech models and posts results to Notion
Meeting notes are available within minutes of the call ending.
Using Fal MCP with Gamut
On Gamut, the Fal MCP server becomes a tool your agents use inside an automated, event-driven workflow — triggered by an event or a schedule, acting through Fal and your other connected MCP servers, with the connection managed by Gamut rather than a local config file.
Frequently asked questions
What AI models does the Fal MCP server provide access to?
Fal hosts a wide range of generative AI models including image generation (Stable Diffusion, Flux), image editing, upscaling, video generation, and speech models. The MCP server exposes these as callable tools for agents.
Can agents run custom or fine-tuned models through Fal?
Yes. Fal supports deploying and running custom models, so agents can trigger inference on fine-tuned models alongside the pre-built model library.
How does Fal handle long-running inference tasks?
Fal runs models on serverless GPU infrastructure with automatic scaling. For longer tasks, agents can submit a job and poll for results, or use webhooks to receive notifications on completion.
Is Fal suitable for production workloads?
Yes. Fal's serverless architecture scales automatically and is designed for production use, with low cold-start times and pay-per-use pricing.