Replicate

Replicate is a cloud platform that makes it easy to run machine learning models with a single API call. Hosts thousands of open-source models including Stable Diffusion, Llama, and Whisper, with pay-per-prediction pricing and fine-tuning support.

Founded in 2019 by Docker and Spotify veterans. Created Cog, an open-source tool for packaging ML models. Grew rapidly with the Stable Diffusion wave in 2022. Offers the largest catalog of deployable ML models via simple API.