AtlasFlux API rate limits and 429 responses - AtlasFlux API

Rate limit tiers

Key Environment	Requests per Minute
Test (`af_test_`)	20
Live (`af_live_`)	60

Generation keys are also protected against excessive concurrent requests. A concurrency rejection uses the same 429 rate_limit_exceeded response and is safe to retry after an active request completes.

Rate limit headers

Every API response includes rate limit headers:

Header	Description
`x-ratelimit-limit`	Maximum requests in the current window
`x-ratelimit-remaining`	Remaining requests
`x-ratelimit-reset`	Unix timestamp when the window resets

Rate limit exceeded

When the rate limit is exceeded, the API returns:

{
  "type": "error",
  "error": {
    "message": "AtlasFlux request rate limit exceeded.",
    "type": "rate_limit_error",
    "code": "rate_limit_exceeded",
    "param": null,
    "request_id": "req_abc123",
    "guidance": "Wait for the retry interval, then send the request again. Use the Retry-After header when available.",
    "action_url": null,
    "retryable": true,
    "retry_after_seconds": 12
  }
}

HTTP status: 429

Best practices

Implement exponential backoff for retries
Respect the x-ratelimit-reset header
Use the idempotency-key header for non-streaming requests to avoid double charges
Use test keys (af_test_) during development