Rate limits

The Partner API uses KV fixed-window limits. A throttled call returns:

429 { "error": "rate_limited" }

with a Retry-After header (seconds). Honor Retry-After and back off.

Limits

Bucket	Limit	Window	Keyed on
All v1 calls	300	60s	your org
Create request	100	3600s	your org
Per-subject creates	5	3600s	subject (across all partners)
Mint viewing session	60	600s	your org

The per-subject limit is shared across all partners

Repeatedly requesting from the same person is throttled regardless of which key you use — it's keyed on the subject, not your org. Design retry logic so you don't burn a subject's budget re-sending the same ask.

Handling 429s

• Read the Retry-After header and wait at least that long before retrying.
• Apply jitter so many workers don't retry in lockstep.
• Treat request-creation throttling separately from read throttling — your polling loop and your create loop hit different buckets.

Next: Errors →