Rate limits
Per-request caps, timing expectations and how to be a good citizen.
There is no global request quota today. Limits are per request and per product, and they exist mostly because the operations behind them are slow.
Per-request caps#
| Endpoint | Cap |
|---|---|
| Mail Verifier | 10 addresses per request (emails) |
| QR Decoder | 8 MB per remote image |
| QR Encoder | size from 1 to 20 modules per pixel |
| YouTube Search (retired) | 50 results per query |
Requests over a cap are answered with a 200 and a message (Mail Verifier) or a 4xx status; they are not queued.
Timing#
Some endpoints do real work per call. Plan for these response times and set client timeouts accordingly:
- Screenshot endpoints: page load plus
delayseconds (default 5), typically 8 to 20 seconds. - Mail Verifier: one SMTP conversation per address, up to ~40 seconds for ten addresses.
- Video Downloader and YouTube Search (both retired): several seconds per call while they remain reachable.
Do not fire these in tight loops from the browser; batch them on your backend.
Keys and volume#
Keyed endpoints are issued for a stated use and expected volume. If your usage grows well beyond what you asked for, tell support so capacity can be planned rather than throttled. The Mail Verifier key in particular carries a per-request address limit that support can raise.
Caching what you get back#
- QR images stay at their returned URL; there is no need to regenerate the same content.
- Screenshot and PDF captures stay at their returned URL as well.
- Direct media URLs from the Video Downloader are short-lived and must be fetched promptly.