Rate limits
One limit, the same for every plan.
100 requests/min, the same limit for every plan — there is no per-plan rate tier and no separate "concurrent renders" cap enforced by the API. Batch your work instead of parallelizing individual calls: a POST /renders call with 100 recipients is one request against this limit, not 100.
A 429 RATE_LIMITED response includes Retry-After — wait and retry rather than backing off blindly.