---
title: Replicate
summary: Run AI models via Replicate's cloud API: image generation, video generation, audio, and thousands of community models.
---

# Replicate

Run AI models via Replicate's cloud API: image generation, video generation, audio, and thousands of community models.

## What it does

Connects an agent to Replicate. Agents can run a model, create a prediction, check a prediction's status, and list available models, covering image generation (Flux, SDXL, Stable Diffusion), video, audio, and thousands of community and custom models.

## Setup

1. Get an API token from your Replicate account settings.
2. Go to **Settings → Integrations → Replicate** and paste in the token.
3. Assign the connection to an agent.

## Credentials needed

A Replicate API token. It's encrypted at rest and never shown again after saving.

## Typical use

An agent that generates an image to accompany a piece of content it's drafting, or one that runs a specialized community model for a task the built-in tools don't cover.

## Related

- [Integrations Guide](/en/docs/guides/integrations-guide): the full catalog and security model these connections share
- [Stability AI](/en/docs/integrations/stability-ai): a similar image-generation alternative