Understanding Claude API rate limits
Rate limits keep the gateway stable and your balance safe. Handling them well means smoother tools and no wasted spend.
·
Traffic limits and spending guardrails
apiToken.sale does not publish a fixed RPM table. A 429 can represent a gateway or upstream capacity limit. The dashboard does not configure request throughput: its per-key guardrails are an optional lifetime spending limit and expiration date.
See also: Claude API best practices
Handling a 429
- Respect the Retry-After header and back off exponentially.
- Reduce concurrency rather than hammering the endpoint.
- Contact support if you need sustained higher throughput.
Frequently asked questions
What are the Claude API rate limits?
apiToken.sale does not publish a fixed RPM number. If you receive a 429, honor Retry-After, back off and reduce concurrency; contact support when you need sustained higher throughput.
What should I do on a 429?
Respect Retry-After, back off, and reduce concurrency; contact support for sustained higher limits.
Use Google or GitHub to create your key and get $10 of Claude usage before you top up.