Koe

• fix(Windows): window controls not working

• fix(mac): unable to download large-v3 whisper model
• add: find and replace in transcript
• add: merge / delete transcript segments
• add: set the time for a transcript segment

• Fix: pressing space when editing the transcript should not play/pause the audio
• Improve whisper model downloading (whisper large-v3 now natively supports Apple Silicon)
• Added new speech-to-text provider: Speechmatics

• Fix: some people are unable to download whisper models due to a bug

If you've previously downloaded any whisper model, they will be ignored and deleted in this new version since we're switching to smaller models in this release, which means you'll need to download the model again

• Support whisper v3 on mac (slightly faster, better result)
• You can now select language and model before transcribing
• Allow to set "prompt" for Local Whisper and OpenAI Whisper
• Allow to use a dedicated provider for dictation
• Show estimated cost for translating the transcript
• Allow to collapse sidebar
• Other minor UI changes
• Various bug fixes

• Fix the app updater, if you can't install the update in the app please download the new version directly!

• Fix: Sometimes the video and audio do not synchronize properly.
• Fix: Properly translate the transcript to Chinese

• redesign the media player
• allow to edit transcript
• allow to hide transcript
• allow to rename
• allow to switch input device in dictation window
• early access feature: generate summary for transcript

• Fix(mac): menubar compatibility issue with the Bartender app

New:

• Windows support
• Local Whisper models support
• Translation support (via OpenAI GPT-3.5 for now)

Redesigned the settings panel, the next release will include support for local Whisper models.

• New: You can now set audio language for each provider separately, because OpenAI and Deepgram supports a different set of languages.
• Fix: Update the languages for OpenAI, which requires ISO-639-1 language code

• Fix: code signing

• New: Add dictation support

• Minor bug fix

• Minor bug fixes

• New: Allow to switch provider and change audio language in sidebar
• New: Add play button to the context menu of transcript segments
• Fix: Main windows is accidentally resize when opening settings window

• Fix(deepgram): reduce length of each sentence in the transcript to make it more readable

• Fix: Export as SRT file, fixed the timestamp format
• New: Export as VTT file

• Ask user to enable microphone access

• Show audio preview after recording
• Fix code signing and notarization issue

Minor bug fix.

Initial public release.

This is a test release note!

• Feature a
• Feature b