Added an "Open Source Licenses" screen under Settings → About. It lists all third-party models and libraries used in the app (Whisper, whisper.cpp, Qwen3.5, Gemma 4, llama.cpp, sherpa-onnx, pyannote segmentation-3.0, 3D-Speaker ERes2Net), each showing the license type, copyright notice, description, and a link to the source.

Apple speech now can utilize diarization

Speaker diarization is now available for local transcripts Chat with your transcripts using local AI New local assistant model support for Gemma 4 and Qwen 3.5 Faster ways to generate summaries, action items, and key takeaways

improve UI for ios 26, add Apple Speech.

Improve performance. Add Core ML support for iOS.

You can now edit transcription text directly from the transcription screen — tap Edit to modify text, then tap "Done Editing" and Save to persist changes. Improved UX for audio export/share and re-transcription flows. General documentation cleanup (internal).