Bring audio in
Record directly into the Media Library or add existing audio files.
Local audio notebook for speaking into writing
Don’t transcribe everything.
Keep the part that matters.
Record or add audio, listen with a waveform, mark useful moments, select one range, and turn only what you need into TXT or SRT.
Listen first.
Textize second.
CowriteAudio is built around deciding what is worth keeping, not assuming every minute needs to become text.
Use the waveform
Play, pause, seek, and let waveform caching keep long previews responsive.
Keep the moment
Add markers and comments, then store one useful range per audio file.
TXT / SRT
Run now or schedule later; SRT can preserve original timeline timecodes for selected ranges.
Tested on an ordinary 8GB business laptop.
One measured test used a 42-minute, 229MB WAV file at 48kHz / 16-bit / mono with the small model. Results vary by hardware and model.
TXT only9:55for the 42-minute test file
TXT + SRT11:30about 27% of source duration in this test
Local-first, with one practical caveat.
Audio processing is local and CowriteAudio does not intentionally upload your audio. Some speech models or dependencies may require a first-time download or local cache setup.
TelemetryNone
AccountNot required
AdminNot required
Auto-updateNone
LanguagesJA / EN / ZH / KO
Shakado 06
Turn the useful part of a recording into something you can write with.
$0+
Windows 10 / 11 · x64 · small model recommended for 8GB RAM PCs