Different strengths and availability
Apple Speech integrates closely with iOS and its on-device language assets. Whisper provides broad multilingual coverage through downloaded local models. Parakeet offers a fast multilingual path for its supported language set. Exact performance depends on the recording and device.
- Apple Speech: system-integrated availability
- Whisper: broad multilingual model coverage
- Parakeet: fast supported-language transcription
- Automatic: resolves a compatible local option
Why the language picker matters
A model should not appear for a language it cannot transcribe. SonWeave shares one catalog between the app and its share extension and filters options according to language support and device capability.
Models may need preparation
The first use of a local model or Apple speech asset can require downloading or initialization. SonWeave displays preparation progress rather than making a new recording look frozen.
FAQ
Frequently asked questions
Which engine should I choose?
Automatic is a practical starting point. Manual selection is useful when you want to compare speed or accuracy for a specific recording.
Do all three engines support 103 languages?
No. The combined app catalog contains 103 options; each engine supports a different subset.
Are the models processed on device?
SonWeave uses local configurations for these engines. Required model or language downloads still need internet access.