1.2.5 - March 15

Highlights

  • Extraction cost and duration tracking
  • New SDK events: silence and live_waveform
  • Extraction context support

Changes

  • Extractions now report cost and duration in the API response
  • SDK: added silence event, fired when no speech is detected
  • SDK: added live_waveform event for real-time audio level visualization
  • SDK: added set_extraction_context() to pass additional context to the extraction engine
  • Recorder now stops automatically when the connection is lost