Serverless Inference

Serverless inference runs models from the Hyperstack AI Studio catalog on demand, with no GPUs to deploy or manage, and you pay only for what you use. Try models interactively in the Text and Image Playgrounds, with real-time parameter tuning and side-by-side comparison, or call them from your own code through the OpenAI-compatible API. To run one model on GPUs reserved for you instead, with its own private endpoint, see Dedicated Inference.

Topics

Text Playground

Run inference with text-to-text models in a chat-style interface, with system prompts, real-time parameter tuning, side-by-side comparison, and API requests.

Image Playground

Generate and edit images with text-to-image and image-to-image models, with real-time parameter tuning and side-by-side model comparison.

Tool Calling

Let a model call functions in your application through the chat completions API, then answer using the results.

Audit Logs

Review a record of every third-party inference request you have made, with its model, latency, token usage, and status.