Rate Limits

Request Rate Limits

Every API endpoint is rate-limited per bearer token (or per IP for unauthenticated requests). Limits reset every 60 seconds.

When you exceed the limit, the API returns:

{
  "statusCode": 429,
  "error": "Too Many Requests",
  "message": "Rate limit exceeded, retry in 42 seconds"
}

The response includes a Retry-After header with the number of seconds to wait.

Plan-Based Usage Limits

In addition to request rate limits, each plan has monthly operation caps and concurrency limits:

Monthly usage

Each call to /upload, /convert, or /extract counts as one operation. Usage resets on the first of each month. When you hit your limit:

{
  "error": "Monthly usage limit reached (50). Upgrade your plan for more."
}

Concurrent jobs

This limits how many jobs can be in queued or processing state at the same time. If you hit the limit:

{
  "error": "Too many active jobs (3/3). Wait for current jobs to finish."
}

Best Practices

• •Reuse bearer tokens — they last 1 hour. Don't call /authorize before every request.
• •Poll with backoff — when checking job status, wait 1-2 seconds between polls instead of hammering /job/:id.
• •Batch uploads first — upload all files, then submit conversion/extraction jobs for each.
• •Handle 429 gracefully — check the Retry-After header and wait before retrying.