Cloud transcription and retry
Proset separates audio capture from transcription. It saves the recording first, uploads it, and then asks a configured cloud service to produce editable text.
Why transcription happens after saving
Saving the audio before transcription creates a recoverable source. If the connection or transcription service is interrupted, Proset can try again without asking you to repeat the idea.
The current web and Android releases do not create an on-device or offline transcript preview. An internet connection is required for the cloud transcript.
The processing sequence
- You finish a recording.
- Proset saves and uploads the audio.
- Cloud transcription enters its queued or processing state.
- The completed text is attached to the saved recording.
The recording detail can open before processing finishes, so seeing a temporary processing state does not mean the audio was lost.
Why retries are safe
Retry uses the saved audio as the same source. Proset can attempt one immediate recovery when you reopen a retryable failure, and the recording detail provides a manual retry action if processing still does not complete.
Proset may route a request through configured primary and backup transcription providers. Provider and model choices can change without changing the user workflow.
Failures caused by monthly limits, spending caps, or unusable audio require a changed condition rather than repeated network retries.
For recovery steps, see Retry a transcription. For the exact states and error responses, see Transcription states and errors.