{ } API & WEBHOOK

Make peace with rate limits

Plan request pacing, retry-backoff and batch sizes so 429 stops being a problem.

Updated:

Rate limits protect the system; they are not the enemy. Used correctly, your integration becomes more reliable and overall Buaze performance stays healthy.

A high 429 ratio is not a performance issue; it is a design issue. Writing with the limit in mind is easier than reacting to it.

Typical limits

Most endpoints have a moderate per-minute ceiling. Costly endpoints like AI Co-Pilot have a lower ceiling, keyed at workspace owner level so a single team member cannot exhaust the quota alone.

Healthy request pattern

  • Add 100-300 ms between sequential requests.
  • Split bulk jobs into batches of 50.
  • Cache: do not re-request the same data within 5 minutes.
  • Backoff: on 429, wait 2-4-8 seconds before retry.
  • Stop after 3 retries and raise an alarm.

Operational monitoring

Track 429 ratio on your integration dashboard. A high ratio is a design issue, not a performance one; revisit your request architecture.

Kontrol listesi / Checklist

  • 429 ratio is tracked.
  • Backoff strategy is implemented.
  • Batch size is defined.
  • Cache TTL is set.
  • Alarm fires after 3+ retries.

SSS / FAQ

Will I lose data on 429?

No. A correct retry mechanism re-sends the request. Data loss is a design flaw, not a consequence of the limit.

Can the limit be raised?

For specific business scenarios, talk to support. First optimize the request pattern.

Did this not solve it?

Write to our support team — we reply within 2 hours. Our median reply time is 12 minutes.

Contact support