Rate Limits
Limits are per key, and per client address for the shared demo key — so one noisy demo user cannot exhaust it for everyone.
Exceeding the limit returns 429 with a Retry-After header in seconds:
{ "error": { "code": "rate_limited", "message": "Too many requests. Slow down and retry.", "retry_after_ms": 4200 }}Back off for Retry-After rather than retrying immediately. The MCP endpoint shares these buckets, so switching transport does not double your budget.