The cheapest machine learning inference

Power your apps with fast AI running on serverless GPUs for incredibly low prices.

Get 1 hour of usage, for free!

Super cheap

Up to $0.00052/second but on average $0.0002/second on A100 GPUs (40GB). Idle time is free.

Hyper optimized

Our team of machine learning engineers performs optimizations under the hood to reduce costs.

Vast model choices

Run popular models like stable diffusion, whisper, and others.

Easy to use

Perform model inference with only 4 lines of code.