Glossary · Data & retrieval

Autothrottle

A crawler pacing itself from the site’s own responses — slowing when latency rises or a 429 arrives, relaxing as it recovers.

What is autothrottle?

Autothrottle learns a per-domain delay from observed response latency and adjusts it dynamically: it doubles the delay (or honours a Retry-After header) when the site answers 429 or 403, and eases back as responses normalize. It is on by default.

Why it matters

A fixed delay is either too slow for a healthy site or too fast for a struggling one. Adapting to the target is both more polite and more effective — a crawl that backs off finishes; a crawl that hammers gets blocked and is refunded nothing for the pages it already read.

Related terms

See it in context

Back to the full glossary

Put your agent to work. Keep the decision.

Give your agent a real browser, authorize the accounts it may use, and keep control of every step. Repeated runs compile into skills that replay without a model.