Skip to content
Docs
Console

Page and section titles. Use your browser's find for words inside a page.

Build with your models.

Boardwalk is an elastic inference platform. Deploy any model from Hugging Face, 8B to 1T, behind one OpenAI-compatible API. It scales from zero in seconds and you pay for the GPU time it holds. Or call an open model from the catalog per token.

Get started

Make your first requestGet an API key and start with Python, TypeScript, or cURL.
Deploy your modelTurn a Hugging Face checkpoint or adapter into a private API.

Keep building

Choose a modelFind model IDs, capabilities, and exact revisions.
How serving worksScale from zero in seconds, one organization per card, exact weights.
Understand pricingGPU time for your models, per-token rates for open models, from prepaid credits.
Explore the APILook up endpoints, request parameters, and errors.

Use an existing client

For OpenAI Chat Completions or Responses, set the base URL to https://api.boardwalk.cloud/v1, supply your Boardwalk key, and choose a model ID.

The Anthropic Messages client is also supported on eligible models. Check each model’s supported_endpoints before choosing a protocol.