Skip to main content

​
Expected Latencies

ConfigurationTypical Latency
Fast75-150ms
Standard400-1400ms
Use extraction_mode=fast and/or resolution_mode=fast to enable fast mode.

​
If Consistently Slow

API latency exceeds expectations

  1. Check network latency - API is hosted on Fly.io (global regions)
  2. Cache responses - Same queries return same results
  3. Use fast mode - Set extraction_mode=fast and resolution_mode=fast for lowest latency
  4. Check your message length - Longer messages take more time to extract keywords from