The hourly request limit is used up

HTTP status 429

This token has used every request it is allowed this hour. Try again after the number of seconds in the Retry-After header.

From Pro up

Each token may send a set number of requests in each clock hour, and this one has sent them all. The count starts again on the hour, and the Retry-After header says how many seconds away that is.

Why you get it

Each token has its own count, kept for each clock hour in UTC. The plan sets how many requests a token may make in an hour, and the prices page lists the number for each plan.

• Two scripts that share one token share its count.
• The /metrics endpoint has its own hourly count, separate from the API's.
• MCP tool calls made with a token use that token's API count.
• A request sent after the count is used up is counted too, and refused, so retrying before the hour turns gains nothing.

Keep an eye on the count

Every answer to a request that was counted carries three headers. RateLimit-Limit is the token's allowance for the hour, RateLimit-Remaining is what is left of it, and RateLimit-Reset is the number of seconds until the count starts again.

What to do

1. Read Retry-After, wait that many seconds, and send the request again.
2. Ask for larger pages with limit, up to 500 rows, so fewer requests fetch the same data.
3. Give each script its own token, so one busy script does not use up another's hour.

Published 27 October 2026 · Checked 27 October 2026