Replicate
Run and deploy open ML models through one API.
About
Replicate catalogs open models packaged to run behind one hosted API, with a browser playground for each. Listings show run counts, cost per run, and hardware requirements. It serves developers who want to test or ship an open model without managing GPU infrastructure.
Features & use cases
- Run open ML models through one hosted API
- Browser playground for testing each model
- Run counts and cost per run on listings
- Hardware requirements listed per model
- Ship an open model without managing GPUs
- Deploy your own models behind the same API
