How to integrate Replicate MCP with Codex

Replicate logo
Codex logo
divider

Introduction

Codex is one of the most popular coding harnesses out there. And MCP makes the experience even better. With Replicate MCP integration, you can draft, triage, summarise emails, and much more, all without leaving the terminal or the app, whichever you prefer.

Also integrate Replicate with

Why use Composio?

Apart from a managed and hosted MCP server, you will get:

  • CodeAct: A dedicated workbench that allows GPT to write its code to handle complex tool chaining. Reduces to-and-fro with LLMs for frequent tool calling.
  • Large tool responses: Handle them to minimise context rot.
  • Dynamic just-in-time access to 20,000 tools across 870+ other Apps for cross-app workflows. It loads the tools you need, so GPTs aren't overwhelmed by tools you don't need.

How to install Replicate MCP in Codex

Run the setup command

Run this command in your terminal to add the Composio MCP server to Codex.

Terminal

It will initiate the authentication in a browser window, authorize Codex to access your Composio account.

Composio authentication page

(Optional) Authenticate with OAuth

To authenticate manually, run the login command to open a browser window and authorize Codex to access your Composio account.

bash
codex mcp login composio

Verify the connection

Run codex mcp list to confirm Composio appears as a registered MCP server.

bash
codex mcp list

Codex App

Codex App follows the same approach as VS Code.

  1. Click ⚙️ on the bottom left → MCP Servers → + Add servers → Streamable HTTP:
  2. Fill the header and Key fields with { "x-consumer-api-key" = "ck_*******" }.
  3. The Key is the Composio API key, that you can find on connect.composio.dev
  4. Click on Authenticate and authorize Codex to your Composio account and you're all set.
Codex App MCP setup
  1. Restart and verify if it's there in .codex/config.toml
bash
[mcp_servers.composio]
url = "https://connect.composio.dev/mcp"
http_headers = { "x-consumer-api-key" = "ck_*******" }

What is the Replicate MCP server, and what's possible with it?

The Replicate MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Replicate account. It provides structured and secure access to your Replicate resources, so your agent can perform actions like running AI model predictions, managing files, browsing model collections, and retrieving model documentation on your behalf.

  • Run and manage AI model predictions: Easily instruct your agent to create, monitor, and manage predictions on any deployed Replicate model using custom input parameters.
  • Browse and discover model collections: Ask your agent to fetch and list available model collections or retrieve example predictions to explore what’s possible on Replicate.
  • Upload and organize files: Let your agent upload new files, list all stored files, or inspect file details to streamline your model workflows.
  • Access model metadata and documentation: Retrieve full model details, schemas, and markdown README docs for any model to help you choose and utilize the right model for your tasks.

Supported Tools & Triggers

Tools
List model collectionsTool to list all collections of models.
Create fileTool to create a file by uploading content.
Create PredictionTool to create a prediction for a given deployment.
Get File DetailsTool to get details of a file by its id.
List FilesTool to list all files created by the user or organization.
List model examplesTool to list example predictions for a specific model.
Get Model DetailsTool to get details of a specific model by owner and name.
Get Model READMETool to get the readme content for a model in markdown format.

Conclusion

You've successfully integrated Replicate with Codex using Composio's MCP server. Now you can interact with Replicate directly from your terminal, VS Code, or the Codex App using natural language commands.

Key benefits of this setup:

  • Seamless integration across CLI, VS Code, and standalone app
  • Natural language commands for Replicate operations
  • Managed authentication through Composio
  • Access to 20,000+ tools across 870+ apps for cross-app workflows
  • CodeAct workbench for complex tool chaining

Next steps:

  • Try asking Codex to perform various Replicate operations
  • Explore cross-app workflows by connecting more toolkits
  • Build automation scripts that leverage Codex's AI capabilities

How to build Replicate MCP Agent with another framework

FAQ

What are the differences in Tool Router MCP and Replicate MCP?

With a standalone Replicate MCP server, the agents and LLMs can only access a fixed set of Replicate tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Replicate and many other apps based on the task at hand, all through a single MCP endpoint.

Can I use Tool Router MCP with Codex?

Yes, you can. Codex fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Replicate tools.

Can I manage the permissions and scopes for Replicate while using Tool Router?

Yes, absolutely. You can configure which Replicate scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do.

How safe is my data with Composio Tool Router?

All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Replicate data and credentials are handled as safely as possible.

Used by agents from

Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai

Never worry about agent reliability

We handle tool reliability, observability, and security so you never have to second-guess an agent action.