Rate limits and usage
Per-minute limits per key, the daily request limit, headers, and the request log.
| Limit | Plan key | Applies to |
|---|---|---|
| Reads per minute, per key | api.requests_per_min (alias api.rpm) | API keys and OAuth apps |
| Writes / sends per minute | half of the read limit | API keys and OAuth apps |
| Requests per day, per organisation | api.requests.day | live API-key and OAuth requests (UTC day) |
A key's own rate_limit.rpm replaces the plan's read figure. Limits are counted across all our servers. Test mode has
its own per-minute counters and no daily limit.
Every response carries RateLimit-Policy, RateLimit and X-RateLimit-Limit/Remaining/Reset. Over a limit you get
429 rate_limited with Retry-After and bucket (read, write, send, ..., or day). The SDKs wait and retry.
Your traffic: GET /v1/api-usage (api_usage_get) for daily and per-key counts and today's remaining daily
requests; GET /v1/request-logs (request_logs_list) for each request of the last 30 days with redacted bodies
(secrets removed, message text replaced by its length, phone numbers and e-mails hidden). Live events:
GET /v1/events/stream (events_stream, Server-Sent Events).