● Documentation · V1 · MAY 2026
Build with GPU.ai.
A REST API, a CLI, and a marketplace of GPU providers — wired through a single binary. These docs cover everything you can do as a developer or as a supplier on the platform.
§ 01 · Start here
JUMP IN
Provision your first instance
From zero to a running H100 in under five minutes.
Quick start →REFERENCEEvery /v1 endpoint
Schemas, examples in curl / Python / Node / Go, response envelopes.
Endpoint reference →FOR PROVIDERSBecome a supplier
List your GPU inventory on the GPU.ai marketplace.
Supplier API →§ 02 · Contents
Getting started
02 chapters
API
04 chapters
- 03ConventionsAuthentication, errors, pagination, idempotency, and rate limits.
- 04Endpoint referenceEvery /v1 endpoint with curl, Python, Node, and Go samples.
- 05Inference APIOpenAI-compatible chat completions and embeddings — no instances to manage.
- 06WebhooksLifecycle events, HMAC signatures, retry policy, verification.
Account
01 chapter
For providers
01 chapter
Need a hand?
We respond fast. Email support@gpu.ai for anything — billing, integration help, weird edge cases.