Documentation
Replicate lets you run AI models with a cloud API, without having to understand machine learning or manage your own infrastructure.
You can run open-source models that other people have published, bring your own training data to create fine-tuned models, or build and publish custom models from scratch.
Get started
Try out Replicate quickly in your environment of choice
Topics
Learn about the building blocks that make Replicate work
A model is a trained, packaged, and published software program that accepts inputs and returns outputs.
Whenever you run a model, you're creating a prediction. Learn about inputs, outputs, files, and other aspects of the prediction lifecycle.
Use deployments for more control over how your models run.
Webhooks provide real-time updates about your predictions.
Organizations let you share access to models, API tokens, billing, dashboards, and more.
Replicate is a pay-as-you-go platform. You are billed for the compute time used to run your models.
The Replicate subprocessor list identifies subprocessors authorized to subprocess customer or personal data on behalf of Replicate to provide our services.
Guides
Ready to build something? Check out these guides to level up your AI skills
Learn how to use Cloudflare and Replicate to build a fast and flexible hosting service for generated images.
Build a Next.js web app that uses Replicate to run models and receive webhooks as they run.
Use Python to build and deploy a Discord chat bot application that uses Flux Schnell via Replicate to generate images from text prompts.
Learn how to build a SwiftUI app that uses Replicate to run a machine learning model.
Package your own custom model using Cog and push it to Replicate as a cloud API.
Learn how to push a Hugging Face Diffusers image generation model to Replicate as a scalable API.
Videos
Prefer to learn by watching videos? Check out some recent demos from our YouTube channel
Run Replicate models using Cloudflare Workers
Create and deploy a web app with a serverless backend and a React frontend in under 60 seconds.
7 minutes
Create stylized videos using pre-trained HuggingFace LoRAs
Make video content using the Hunyuan model with pre-trained styles from HuggingFace, or using your own images as training data.
3 minutes
FLUX.1 Schnell vs FLUX.1 Dev
Explore the differences between Flux Schnell and Flux Dev image generation models and learn how to enhance image quality effectively.
6 minutes
David Attenborough is now narrating my life
Here's a GPT-4-vision + ElevenLabs python script so you can star in your own Planet Earth.
2 minutes
Write your shell commands in English
Use language models like GPT-4o and Llama to write one-liner shell commands, then execute them.
4 minutes
Using webhooks with Replicate's API
Learn how to receive webhooks from Replicate's API when running predictions and trainings.
14 minutes