Skip to content

Add language parameter support for Whisper transcription#384

Merged
HenryNdubuaku merged 1 commit intomainfrom
feature/language-parameter
Feb 22, 2026
Merged

Add language parameter support for Whisper transcription#384
HenryNdubuaku merged 1 commit intomainfrom
feature/language-parameter

Conversation

@rshemet
Copy link
Copy Markdown
Collaborator

@rshemet rshemet commented Feb 21, 2026

Summary

Adds language parameter support for Whisper transcription in both streaming and file modes.

Changes

  • Stream API: cactus_stream_transcribe_start(model, language="es")
  • CLI: cactus transcribe --language es --file audio.wav
  • Cloud fallback: Language passed to cloud API for accurate transcription
  • DRY: Shared prompt builder for consistent language handling

Testing

  • ✅ Live streaming with custom language
  • ✅ File transcription with custom language
  • ✅ Cloud handoff preserves language
  • ✅ Backward compatible (defaults to 'en')

Implementation

  • Non-breaking: Defaults to English
  • Whisper-only: Moonshine gets empty prompt as before
  • 85 net lines changed across 4 files

- Add language option to streaming transcription API with default 'en'
- Pass language to cloud fallback for accurate transcription
- Add --language CLI flag for both file and live transcription
- Use shared prompt builder to maintain DRY principle

Signed-off-by: Roman Shemet <[email protected]>
@HenryNdubuaku HenryNdubuaku merged commit f431494 into main Feb 22, 2026
1 of 2 checks passed
ncylich pushed a commit that referenced this pull request Feb 24, 2026
- Add language option to streaming transcription API with default 'en'
- Pass language to cloud fallback for accurate transcription
- Add --language CLI flag for both file and live transcription
- Use shared prompt builder to maintain DRY principle

Signed-off-by: Roman Shemet <[email protected]>
cattermelon1234 pushed a commit to cattermelon1234/cactus that referenced this pull request Feb 28, 2026
…ute#384)

- Add language option to streaming transcription API with default 'en'
- Pass language to cloud fallback for accurate transcription
- Add --language CLI flag for both file and live transcription
- Use shared prompt builder to maintain DRY principle

Signed-off-by: Roman Shemet <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants