API Rate Limits

The same rate limits apply to every account. Metered usage measures total volume; rate limits control short bursts.

Per-minute buckets

These per-minute limits apply per API key to direct API calls. Use GET /usage to see the active bucket for your key.

Inference

Detection endpoints such as /detect, /detect-url, and /detect-urban.

1000 rpm

OCR

Metadata-strip OCR endpoints.

1000 rpm

Reference

Species, countries, models, and detection retrieval endpoints.

1000 rpm

Usage

GET /usage for usage and quota state polling.

300 rpm

Concurrency

Inference and OCR requests are additionally limited to 20 concurrent requests per API key.

Free tier throttle

Accounts without active API billing are on the free tier (250 detections per month) and are additionally throttled to 30 requests per minute across the whole organisation. Enable API billing to lift the throttle and unlock the full per-key limits above.