New here? The quickstart takes you from an API key to a first response.
What you can call
The base URL ishttps://api.overcontrolgroup.com/v1. You authenticate with an API key from the API dashboard. Chat Completions and Messages share that key and the same models.
Moving a client you already have is a base URL, your key, and a Lume model. Migrate from OpenAI and Migrate from Anthropic show that change.
Both are text APIs. You can generate a reply, stream it as it is produced, ask for JSON, and let the model call functions your application defines. Examples on this site use
lume-3.5.
lume-omni can read an image, an audio clip, or a PDF and answer in text. Messages can send that model an image. Audio and PDF are Chat Completions requests. The other models take text.
Models lists every ID the API serves, with context windows and prices. Pricing explains how tokens are billed, and how the included text requests scale from Free to Plus and Pro. Authentication covers the key, the scopes on it, and the errors you get back when a call is not allowed.
Request logs leave out the prompt and the completion. A Chat Completions request that sets store to true is declined, with HTTP 400 and the message 'store=true' is not supported.
The privacy policy and the terms of service describe how OverControl handles data.
Where to go next
Make your first request
Create an API key and call Chat Completions and Messages.
Choose a model
Compare context, output, input type, and price across the Lume family.
Authenticate
See how keys, headers, and scopes work, and what a rejected key looks like.
Review pricing
Token rates in euros and dollars, and how included requests scale by plan.