Replicate API MCP Server

Cloud PlatformsTypeScript

Run Replicate AI models through a simple tool-based interface

Available Tools

search_models

Find models using semantic search

list_models

Browse available models

get_model

Get details about a specific model

list_collections

Browse model collections

get_collection

Get details about a specific collection

create_prediction

Run a model with your inputs

get_prediction

Check a prediction's status

cancel_prediction

Stop a running prediction

list_predictions

See your recent predictions

view_image

Open an image in your browser

clear_image_cache

Clean up cached images

get_image_cache_stats

Check cache usage

The Replicate API MCP provides a seamless integration with Replicate's model hosting platform, allowing you to search, run, and manage AI models directly from your AI assistant. Access thousands of open-source models including text-to-image generators, language models, and specialized AI tools through a simple interface. With this MCP, you can search for models using semantic search, create and manage predictions, and view generated images directly in your browser. The server handles all the complexity of the Replicate API, providing a streamlined experience for running state-of-the-art AI models without leaving your assistant.

Installation

You can install the Replicate API MCP in several ways:

Global Installation (Recommended)

Install the server globally using npm:

npm install -g mcp-replicate

Run with npx

Alternatively, you can run it directly with npx:

npx mcp-replicate

Install from Source

If you prefer to install from source:

git clone https://github.com/deepfates/mcp-replicate
cd mcp-replicate
npm install
npm run build
npm start

Configuration

The server requires a Replicate API token to function. You can obtain one by:

Going to the Replicate API tokens page
Creating a new token if you don't already have one
Copying the token for configuration

Setting Up in Claude Desktop

Open Claude Desktop Settings (⌘ ,)
Select the "Developer" section in the sidebar
Click "Edit Config" to open the configuration file
Add the following configuration, replacing your_token_here with your actual Replicate API token:

{
  "mcpServers": {
    "replicate": {
      "command": "mcp-replicate",
      "env": {
        "REPLICATE_API_TOKEN": "your_token_here"
      }
    }
  }
}

Using Environment Variables

If you're using another MCP client, you can set the token as an environment variable:

export REPLICATE_API_TOKEN=your_token_here

Usage

Once configured, you'll see a 🔨 hammer icon in the bottom right corner of new chat windows in Claude Desktop, indicating that the tools are available.

You can ask your assistant to:

Search for specific types of models
Run models with your inputs
Check on the status of running predictions
View generated images
Browse model collections

Troubleshooting

Tools Not Appearing

If the tools aren't showing up:

Verify that Claude Desktop is properly configured with the MCP server settings
Ensure your Replicate API token is set correctly
Try restarting both the server and Claude Desktop
Check the server logs for any error messages

Tools Visible But Not Working

If you can see the tools but they're not functioning:

Verify your Replicate API token is valid
Check your internet connection
Look for any error messages in the server output

Requirements

Node.js >= 18.0.0
TypeScript >= 5.0.0
Claude Desktop or another MCP client for using the tools