higgsfield-mcp-server
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@higgsfield-mcp-serverGenerate a photorealistic image of a golden retriever in a spacesuit"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
higgsfield-mcp-server
Remote MCP server that lets Claude generate images and videos through the Higgsfield AI API — photorealistic photos (Soul), reference-guided images (Popcorn), and video (Google Veo 3.1, Kling, ByteDance Seedance) — directly from chat.
Tools exposed
Tool | What it does |
| Text-to-image, photorealistic (Soul/Standard) |
| Text-to-image with up to 8 reference photos (Popcorn/Auto) |
| Text/image/frame/reference-to-video via Google Veo 3.1 |
| Cinematic video via Kling v2.1 / v2.5 |
| Fast/cheap video via ByteDance Seedance |
| Poll a generation job until it's done |
| Cancel a queued job |
| List every available endpoint (incl. ones only reachable via raw_request) |
| Call any other Higgsfield endpoint (Soul character/reference, FLUX Kontext, MiniMax Hailuo, DOP) |
All generation tools are asynchronous: they return a request_id immediately;
call higgsfield_check_status (Claude will do this automatically) until the
job is completed.
Related MCP server: gemini-nano-banana-mcp
1. Get your Higgsfield API keys
Go to https://cloud.higgsfield.ai and sign in.
Create an API key (you did this already if you're reading this after asking Claude for help) — you get an API Key ID and an API Key Secret. Copy both, they're shown only once.
2. Deploy this server (free tier works)
You need somewhere that keeps a Node process running 24/7 and reachable over HTTPS. Two free options that need no credit card for a small server like this:
Option A: Render.com
Push this folder to a new GitHub repository (private is fine).
On https://render.com → New + → Web Service → connect that repo.
Settings:
Runtime: Node
Build Command:
npm installStart Command:
npm start
Under Environment, add these three variables:
HF_API_KEY_ID= your Higgsfield API Key IDHF_API_KEY_SECRET= your Higgsfield API Key SecretMCP_ACCESS_TOKEN= any long random string you make up (e.g. runopenssl rand -hex 32locally, or just mash the keyboard for 40+ characters) — this is the password Claude will use to talk to your server, so nobody else can spend your Higgsfield credits.
Deploy. Render gives you a URL like
https://higgsfield-mcp-server.onrender.com. Your MCP endpoint is that URL +/mcp, e.g.https://higgsfield-mcp-server.onrender.com/mcp.
Note: Render's free web services sleep after inactivity and take a few seconds to wake up on the next request — fine for occasional use.
Option B: Railway.app
Same idea: new project from this GitHub repo, set the same three environment
variables, Railway auto-detects npm start. You get a URL like
https://higgsfield-mcp-server-production.up.railway.app, endpoint is that +
/mcp.
3. Add it to Claude as a custom connector
In Claude (claude.ai / desktop app) → Settings → Connectors → Add custom connector:
Name: Higgsfield
URL:
https://<your-deployed-url>/mcpAuthentication: Bearer token → paste the
MCP_ACCESS_TOKENvalue you set in step 2.
Enable it for your chat, and Claude will be able to call the tools above.
Local testing (optional)
cp .env.example .env # fill in real values
npm install
npm startThe server listens on http://localhost:3000/mcp (or $PORT).
Security notes
HF_API_KEY_ID/HF_API_KEY_SECRETgive full access to your Higgsfield account and billing. They live only in the host's environment variables — never commit them, never put them in the URL.MCP_ACCESS_TOKENis what stops a random person who finds your server URL from generating content on your Higgsfield credits. Treat it like a password.Every generation call spends Higgsfield credits per your plan's pricing.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables AI-powered image and video generation using Higgsfield AI models through MCP-compatible clients like Claude Desktop and Perplexity.352MIT
- AlicenseNot gradedqualityCmaintenanceMCP server that enables Claude Code to generate images using Google's Gemini image generation models.MIT
- AlicenseNot gradedqualityDmaintenanceMCP server for Higgsfield AI that enables image and video generation using 16+ models through Claude, Cursor, or any MCP-compatible client.3523MIT
- AlicenseNot gradedqualityCmaintenanceA local MCP server that enables Claude to generate images using the OpenAI API.9MIT
Related MCP Connectors
MCP server for Grok Imagine AI video generation
MCP server for Hailuo (MiniMax) AI video generation
MCP server for OpenAI Sora AI video generation
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/timursatubaldiyev-creator/higgsfield-mcp-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server