Rate limits
Per-key request limits and per-project concurrency.
Limits are per API key (requests per minute) and per project (concurrent streams). Responses carry x-ratelimit-limit-requests, x-ratelimit-remaining-requests and x-ratelimit-reset-requests; a 429 carries retry-after in seconds.
Default limits: 600 requests per minute per API key, 2,000,000 tokens per minute per project, and 64 concurrent streams per project. Write to support if a workload needs more — for example many agents streaming in parallel.