Skip to main content
Every request is rate-limited to keep the API fast and fair. Limits are applied per API key when you send one, otherwise per IP address. Adding a key raises every limit below.

Budgets

There are two windows — a per-minute burst limit and a per-day quota. Search and raw-source endpoints have their own (lower) per-minute limits because they’re heavier.
The cheapest way to raise your effective throughput isn’t a bigger key — it’s caching. Responses are served from the edge and carry ETags, so well-behaved clients rarely touch the rate limiter at all.

Headers

Every response includes your current standing on the per-minute window:
Requests served from the edge cache don’t count against your budget — only requests that reach the origin do. A cache HIT (see the X-Cache header) is effectively free.

Handling 429

When you exceed a limit you get a 429 with a Retry-After header (in seconds) and the standard error envelope:
Back off and retry after the indicated delay. A simple, correct pattern:
Don’t hammer through a 429 in a tight loop — repeated violations can get an IP temporarily blocked at the edge. Respect Retry-After.