Claude API 429 (rate_limit_error)

Is there an outage right now?

Claude API official status is operational. If this error persists, it is likely an implementation or environment issue on your side.

Updated: Jul 21, 2026, 8:41 PM UTC

View Claude API status

What this error means

429 means you hit a rate limit: RPM/ITPM/OTPM (per-minute input/output tokens) exceeded.

Common causes

  • Per-minute request/token limit exceeded
  • Too much concurrency
  • Low limits on your tier
  • Temporary demand spike on Anthropic side

How to fix

  1. 1.Exponential backoff honoring Retry-After
  2. 2.Lower concurrency and rate
  3. 3.Use prompt caching / batch to cut usage
  4. 4.Check/raise your tier limits
  5. 5.Check status at top of page

When waiting helps (and when it won't)

May clear over time if demand-driven. If it is your own limit, waiting will not help—reduce rate.

Alternatives when Claude API is down

OpenAI APIGemini API / AI StudioMistral APIDeepSeek API

Related errors

FAQ

What is 429 rate_limit_error?

429 means you hit a rate limit: RPM/ITPM/OTPM (per-minute input/output tokens) exceeded.

How do I fix 429 rate_limit_error?

Exponential backoff honoring Retry-After / Lower concurrency and rate / Use prompt caching / batch to cut usage / Check/raise your tier limits / Check status at top of page

Compare Claude API with others