Overview
Reference for Nebula Block's two APIs — the OpenAI-compatible Inference API and the Platform API.
Nebula Block exposes two APIs. They share one credential, so a single API key works against both.
https://inference.nebulablock.com/v1
Running models: chat, vision, images, video, embeddings, reranking
https://api.nebulablock.com/api/v1
Managing resources: instances, SSH keys, API keys, object storage, billing
The Inference API implements the OpenAI API surface, so existing OpenAI SDKs work against it by changing the base URL. The Platform API is a conventional REST API that wraps every response in a data / message / status envelope.
Authenticating
Both APIs use Bearer authentication:
Authorization: Bearer sk-...Create a key in the console under API Keys. See Authentication for access tokens, login, and the rules around key rotation.
Rate limits
Inference calls are subject to per-minute and per-day limits that depend on your tier, plus a per-model daily cap. Exceeding either returns 429. See Tiers and Rate Limits.
Before you provision compute
Renting GPU instances requires an SSH key on your account, so that you can reach the instance once it is running. Add one in the console under SSH Keys, or with Create SSH Key.
See also
Last updated