Replicate (npx) MCP Server
CommunityContributed by Replicate
Run AI models on Replicate. Create predictions, search for models, manage deployments, and generate images, audio, and text.
About the Replicate (npx) MCP Server
The Replicate (npx) MCP server is a local (stdio) Model Context Protocol server available in the McpMux registry. Run AI models on Replicate. Create predictions, search for models, manage deployments, and generate images, audio, and text. This is a community-contributed MCP server by Replicate.
Install the Replicate (npx) MCP server with one click using McpMux. It works with Cursor, Claude Desktop, Claude Code, VS Code, ChatGPT, Windsurf, JetBrains, and any MCP-compatible AI client. This server requires an API key — McpMux securely stores your credentials with AES-256-GCM encryption.
Transport Configuration
{
"type": "stdio",
"command": "npx",
"args": [
"-y",
"replicate-mcp@latest"
],
"env": {
"REPLICATE_API_TOKEN": "${input:REPLICATE_API_TOKEN}"
},
"metadata": {
"inputs": [
{
"id": "REPLICATE_API_TOKEN",
"label": "Replicate API Token",
"description": "API token for the Replicate platform.",
"type": "text",
"required": true,
"secret": true,
"placeholder": "r8_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
"obtain": {
"url": "https://replicate.com/account/api-tokens",
"instructions": "1. Go to replicate.com/account/api-tokens\n2. Click 'Create token'\n3. Name your token and copy it (starts with r8_)",
"button_label": "Create Token"
}
}
]
}
}Supported AI Clients
The Replicate (npx) MCP server works with all MCP-compatible AI clients through McpMux:
Related MCP Servers
Perplexity (npx)
AI-powered search using Perplexity models. Provides web search, deep research, and reasoning capabilities with real-time web context.
Exa Search (Remote)
AI-powered web search that understands meaning, not just keywords. Hosted remote MCP server, no local setup required.
Exa Search (npx)
AI-powered web search that understands meaning, not just keywords. Find similar pages, search by content type, and get clean text from any URL via npx.
ElevenLabs (uvx)
Generate speech, clone voices, create sound effects, and manage audio with the ElevenLabs API. Text-to-speech, voice design, and audio isolation.
Pinecone (npx)
Manage Pinecone vector database indexes and records. Create indexes, upsert vectors, query by semantic similarity, and manage namespaces.
Qdrant (uvx)
Store and retrieve information from Qdrant vector database using semantic search. Supports local and cloud Qdrant instances.
Install Replicate (npx) with McpMux
One-click install from the McpMux desktop app. Auto-configures for Cursor, Claude, VS Code, ChatGPT, Windsurf, JetBrains, and any MCP-compatible client.