Figma MCP Server logo

Figma MCP Server

Connects MCP-compatible assistants to Figma REST data for file analysis, components, exports, comments, and project metadata.

Install Figma MCP Server

Pick your client, then add it in one click or paste the config.

Streamable HTTP (Remote)
Server URL
https://api.figma.com/v1/oauth/token

Remote servers are added in the app, not the config file: Customize → Connectors → Add custom connector, then paste this URL.

Overview

Figma MCP Server MCP server exposes Figma REST API capabilities to AI assistants through MCP over a long-running HTTP service. It runs on Apify as a standby Actor or locally with Node.js and Apify tooling, using the `/mcp` JSON-RPC endpoint. Authentication supports a Figma personal access token or optional OAuth 2.0 for shared team and file access. Reach for it when an agent needs to inspect Figma files, examine components, export design nodes, manage comments, or read project and team metadata.

Use cases

  • Analyze Figma file structure and styles
  • Inspect component details and usage
  • Export Figma nodes into image or document formats
  • Read and create design comments
  • Browse Figma projects and team metadata

Key features

  • Figma REST API access over MCP
  • Depth-limited file analysis
  • Component listing and inspection
  • PNG, SVG, PDF, and JPG node export
  • Comment reading and creation
  • Personal access token and OAuth authentication

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 2d ago

Recent health check succeeded.

7/7 checks healthy over the last 46d

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars1Stargazers on the source repository.
Last commit7mo agoMost recent push to the default branch.
Directory activity2 viewsConfig copies, upvotes, and views on AllMCPs.

How Figma MCP Server works

What Figma MCP Server MCP server does

The Figma MCP Server MCP server gives compatible AI clients an MCP interface to selected Figma REST API operations. It supports file inspection, component discovery, asset export, comments, and project or team metadata. File analysis can return structure, metadata, styles, and depth-limited trees. Component operations cover listing, details, and usage information.

The server also includes MCP resources and prompts intended to guide common Figma workflows. The available tool set includes operations such as analyze_file, get_file_structure, list_components, export_node, get_comments, and create_comment; exact names may vary slightly in the implementation. Design modification helpers are present, but the project notes that REST API coverage may not support deep design writes.

How it works

The Figma MCP Server MCP server exposes MCP JSON-RPC at POST /mcp. On Apify, the endpoint is normally available at the run’s Container URL followed by /mcp. A client sends MCP requests to that URL while the Actor is running. The service also provides / for service information and /health for liveness and authentication status.

Authentication can use a personal access token supplied as figmaAccessToken in Actor input or through FIGMA_ACCESS_TOKEN for local execution. When a token is present, it is used for Figma API calls and OAuth routes are disabled. Alternatively, the server can start an OAuth authorization flow through /oauth/authorize; after consent, Figma redirects to /oauth/callback, and the refreshed session is stored in the Apify key-value store for that run.

Setup and configuration

For Apify deployment, install dependencies, authenticate with the Apify CLI, and run apify push. Start the Actor as a long-running web server, provide at least a Figma token or the required OAuth app fields, and copy the active Container URL into the MCP client configuration. The default local port is 8080, although Apify may override it.

Local development uses npm install followed by apify run. The server can also be started with Node when authentication and input are supplied through the environment or another mechanism. OAuth requires a client ID, client secret, an exact registered redirect URI, and optionally a comma-separated scope list. The documented default scopes include file content and comment read/write access.

Tools and capabilities

The Figma MCP Server MCP server provides capabilities for:

  • Analyzing file structure, metadata, styles, and nested content
  • Listing components and retrieving component details or usage
  • Exporting nodes as PNG, SVG, PDF, or JPG
  • Reading and creating Figma comments
  • Listing projects and teams with related metadata
  • Providing MCP resources and prompts for guided workflows

GET responses can be cached through the enableCaching input, which defaults to true. maxConcurrentRequests is configurable but marked as reserved for future use.

Limitations and notes

Apify Container URLs are tied to individual runs and stop working when those runs end. A new run may therefore require an updated client URL. OAuth stores one session per run in the default key-value store and is not designed for multiple concurrent users in one process.

The Figma MCP Server MCP server uses the REST API rather than the Figma Plugin API. As a result, operations requiring deeper design modification may need a separate Plugin API approach. A personal access token also takes precedence over OAuth: setting one disables the OAuth endpoints for API authentication.

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Figma MCP Server

Figma MCP Server is an MCP server that connects AI assistants to Figma through the Figma REST API. Its main tools analyze files, inspect components, export nodes, read and create comments, and access project or team metadata, with MCP resources and prompts available for guided workflows.

Install it with `npm install`, then run `apify run` locally or use `apify login` and `apify push` for an Apify deployment. Configure a Figma personal access token or OAuth app settings, and connect the client to the running `/mcp` endpoint.

Yes, Figma MCP Server works with Cursor through its remote HTTP MCP configuration. Deploy and start the Actor, place the run’s Container URL followed by `/mcp` in Cursor’s `mcp.json`, and restart Cursor after changing the URL.

Use a Figma personal access token by setting `figmaAccessToken` in Actor input or `FIGMA_ACCESS_TOKEN` for local execution. OAuth 2.0 is also available when no personal access token is set, using the authorization and callback routes with a configured Figma OAuth app.

Related MCP servers

Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewFigma MCP Server AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/figma-mcp-server-2?style=directory)](https://allmcps.com/mcp/figma-mcp-server-2)
HTML Embed
<a href="https://allmcps.com/mcp/figma-mcp-server-2"><img src="https://allmcps.com/api/badge/figma-mcp-server-2?style=directory" alt="Figma MCP Server on AllMCPs" /></a>