Rate limits

How many requests you can make, how the limit is counted and how to back off.

Surface	Limit	Counted per
Partner API `/api/partner/v1/*`	600 requests / minute	Seller connection (grant): a busy seller never slows another one down.
Token endpoint `/api/oauth/token`	60 requests / minute	Client id + caller IP address.

Fixed one-minute windows. Every answer carries the standard headers so you can slow down before hitting the ceiling:

Headers

RateLimit-Limit: 600
RateLimit-Remaining: 587
RateLimit-Reset: 42

RateLimit-Reset is the number of seconds until the window resets. Over the limit you get:

429 Too Many Requests

{
  "error": {
    "message": "Too many requests. Slow down and retry after the window resets.",
    "code": "RATE_LIMITED"
  }
}

Wait RateLimit-Reset seconds, then retry. Add jitter when many workers share one connection.
Use webhooks instead of polling, and updatedSince for read-backs.
Page with limit=200 for bulk syncs: fewer, bigger pages.