Rate limits

Caps reset on a rolling window. Both apply — the lower one binds:

Per-route limits also apply on top of these. Bursting against POST /api/postsstill respects that route's own per-key bucket (10 posts / min by default).

Apply to the Jestha account behind the key, not the key itself. Posting under multiple keys owned by the same account draws against the same bucket.

When media is uploaded through an API key, tighter limits apply than for direct UI uploads:

Every response to an API-key-authenticated request carries these headers — including 2xx, 429, and most 4xx. Use them to throttle proactively instead of waiting for a 429.

X-RateLimit-Limit:     60                # per-minute cap
X-RateLimit-Remaining: 42                # requests left in the current minute
X-RateLimit-Reset:     1716889389        # unix timestamp when the minute window resets
X-Jestha-Api-Version:  1                 # current API major version

When you exceed a limit, the response is:

HTTP/1.1 429 Too Many Requests
Content-Type: application/json

{
  "error": "Rate limit exceeded",
  "message": "This API key is limited to 60 requests per minute.",
  "retryAfter": 17
}

retryAfteris the number of seconds until the window resets. Back off and try again then — don't hot-loop, which can extend the lockout if it looks like an attack.

The 1,000 posts/day quota for content via API keys is per-account. When hit:

HTTP/1.1 429 Too Many Requests
Content-Type: application/json

{
  "error": "Daily content quota exceeded",
  "message": "Your account has hit the 1000 posts/day limit for content posted via API keys. Try again tomorrow.",
  "retryAfter": 72345
}

The reset is at UTC midnight. If you genuinely need a higher cap, get in touch at [email protected].

• Cache responses where the data doesn't change minute-to-minute.
• Subscribe to webhooks instead of polling — they're free of rate limits and lower latency.
• Honour retryAfter exactly. Polling sooner risks an extension.
• If you expect bursty volume (e.g. you cross-post in batches), spread the writes across the minute window rather than firing them in one tight loop.