Limits
Per-key limits on the Ablatic API and how to raise them
Limits are enforced per API key. This page lists the default limits and how to request more capacity.
Default limits
Each new account is provisioned on the api billing model with these limits (you create your one key after first login):
- 20 requests per minute
- 5 concurrent requests
- 2 agent slots
- EUR 50 monthly budget
- 1 active key
- expires 21 days after provisioning
Agent slots gate the server-side agent loop. With zero slots, an agent run is refused.
A request may carry up to 2000 messages across /v1/chat/completions, /v1/messages, and /v1/responses; the effective limit is the 200,000-token context window, so long agentic loops are bounded by context, not message count.
Raising your limits
There are no fixed packages. We set every limit per account when we provision it, and you can request a higher value for any of them: requests per minute, concurrent requests, agent slots, monthly budget, and the number of keys.
Send a capacity request from the console, describing what you need and why. An admin reviews it and applies the approved values to your key, and you get an email once it is done.