1. Fixed potential shortcut key conflicts 2. Fixed issues where the microphone may suddenly malfunction 3. Fixed other bugs 4. Updated the speech recognition framework to the latest version

1. Added IAP Pro version, welcome to support Flow! 2. Fixed an issue where audio might continue to be captured after releasing the hotkey. 3. Fixed other known issues.

Better looking permission request window

Better-looking permission authorization UI

1. Add ew models: Qwen3-ASR, FireRedASR 2. Add ESC key to stop recording 3. Update the voice recognition framework to the latest version

1. Added support for FunASR Nano 2. Add FunASR Nano INT8 quantitative version model to the built-in model 3. Update the speech recognition framework to the latest version 4. Fix known issues

1. Added microphone test function 2. Improve the recognition speed of Apple’s built-in models on macOS26 3. Optimize audio processing logic 4. Fixed the problem of being unable to input after the recognition is completed. 5. Selecting a microphone now does not change system settings 6. Optimize real-time recognition effect

AI correction can add multiple prompts and switch quickly.

1. add status menu quickly select models function

1. Adjust the UI of Apple's built-in model 2. Update the speech recognition engine to the latest version

add macOS26 Apple’s on-device voice model

1. Add Traditional Chinese, Arabic, Czech, Danish, Dutch, French, German, Greek, Hebrew, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Thai, Turkish, Ukrainian, Vietnamese 2. Fix minor bugs in UI

1. Add AI correction feature 2. Add push-to-talk keyboard shortcut 3. Add Nvidia Parakeet Tdt 0.6B V3 INT8 model 4. Upgrad speech recognition framework to the latest version 5. Optimize model icon display 6. Fix issue where status bar shortcut keys were not displaying correctly 7. Optimize UI on some pages 8. Fix bug that might cause crashes

1. Model page UI beautification

1. Fixed the volume visualization in the floating panel, which can now be automaticly adjusted according to the volume.

1. Put the Shortcut settings on a separate page 2. Add SenseVoice INT8 quantized version 3. Add HotkeyManager acknowledgment page

1. Now you can set single keyboard as shortcut, and support fn and other keys! 2. Added sound effects, sound effects can now be played at the beginning and end of recognition. 3. Voice recognition framework updated. 4. Fixed the issue where the floating panel could not be closed normally.