Replicate makes it easy to run machine learning models in the cloud from your own code.

You can run open-source models, or run your own public and private models.

Want to see examples of what you can build with Replicate? Check out our showcase.

Run models

Run a model from Node.js - Get started with a few lines of JavaScript Run a model from Python - The lingua franca of the machine learning world Run a model from Google Colab - Cloud-hosted Jupyter notebooks Build a website with Next.js - React + Node.js for rapid development Build an app with SwiftUI - Develop for macOS, iOS, and (soon) visionOS Build a Discord bot - Chat-based image generation Build an app with Elixir - A high-performance successor to Ruby

Push models

Push a model to Replicate - Use Cog to build and push your own models Push a private model - Collaborate privately with your team Fine-tune an image model - Train a new model on faces or styles Fine-tune a language model - Train a new model using your private data Get a GPU machine - Access powerful cloud compute on Lambda Labs Push a Diffusers model - Stable Diffusion and countless others Push a Transformers model - Attention is all you need Push a model using GitHub Actions - Continuous automated model deployment

Learn more

How does Replicate work? - A guide to core concepts Showcase - Imagine what you can build Using Webhooks - Get realtime updates about your predictions Client libraries - JavaScript, Python, Ruby, Swift, Elixir HTTP API reference