Rate Limits

The API rate-limits each request to keep the service fast and fair. Exceeding a limit returns 429 Too Many Requests.

Per-token limits#

Limits are tracked per API token — separate from the per-IP limits applied to unauthenticated traffic. Each token gets its own bucket, so one integration's usage never starves another's.

Limits by endpoint#

Response headers#

Successful responses carry your current standing, so you can pace yourself instead of discovering the limit by hitting it. So do 429s. Other errors do not.

Limits are counted per bucket, and your bucket is your API token — so these numbers describe that token's usage, not your workspace's total.

429 responses#

When you hit a limit, the response carries a Retry-After header (seconds to wait). Always honor it rather than retrying immediately.

Best practices#

Use jittered exponential backoff on 429 (and on 5xx):

async function withBackoff(fn, max = 5) {
  for (let attempt = 0; ; attempt++) {
    const res = await fn();
    if (res.status !== 429) return res;
    if (attempt >= max) return res;
    const retryAfter = Number(res.headers.get("Retry-After")) || 2 ** attempt;
    const jitter = Math.random() * 0.3 * retryAfter;
    await new Promise((r) => setTimeout(r, (retryAfter + jitter) * 1000));
  }
}

The Rate Limits & Backoff guide goes deeper, with samples in all languages.