# Replicate

> Use this tool when you need to integrate Replicate's hosted AI models into your application through a standardized Model Context Protocol (MCP) interface. It solves problems of accessing and managing diverse machine learning models by providing a bridge to Replicate's platform, supporting operations like model listing, prediction creation, and metadata access. With inputs like model requests and outputs like prediction results, use it when building MCP-compatible applications that require easy access to pre-trained models for tasks like image generation and text processing.

Canonical page: https://skillsregistry.net/skills/deepfates-replicate  
JSON: https://api.skillsregistry.net/v1/skills/deepfates-replicate

## Description

This Replicate MCP server implementation provides a bridge between the Model Context Protocol and Replicate's AI model hosting platform. Developed as an open-source project, it enables seamless interaction with Replicate's diverse collection of machine learning models through a standardized MCP interface. The server supports key operations such as listing available models, creating and managing predictions, and accessing model metadata. Built with TypeScript and leveraging the official Replicate API, it offers robust error handling, caching mechanisms, and webhook support. This implementation is particularly useful for developers and researchers looking to integrate Replicate's hosted AI models into their MCP-compatible applications, enabling easy access to a wide range of pre-trained models for tasks like image generation, text processing, and more.

## Trust

- **Trust score (0–1):** 0.50
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** media
- **Updated:** 2026-05-21

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/deepfates-replicate)
- **Repository:** <https://github.com/deepfates/mcp-replicate>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "deepfates-replicate"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/deepfates-replicate` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/deepfates-replicate/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt