Shakado 06 — CowriteAudio A local audio notebook for recording, listening, marking, and turning the useful parts of your voice and audio files into text. CowriteAudio is not only a transcription tool. It is a small local workspace for people who think by speaking. Why it exists Most transcription tools start from one assumption: put in an audio file and get the full text back. CowriteAudio is built for a different workflow. Sometimes you do not need the whole recording. You need the moment where the idea changed, the phrase worth keeping, or the range you want to turn into writing, notes, or subtitle material. CowriteAudio lets you record or add audio, look at a lightweight waveform, listen, add markers and comments, select one useful range, and textize only what you need. 30 seconds to use 1. Launch CowriteAudio and add an audio file, or record a new one inside the app. 2. Listen with the waveform, add markers or comments, and set the range you want to use. 3. Run textization now or schedule it for later, then save the result as TXT and, when needed, SRT. What it does - Record audio directly into the Media Library. - Add existing audio files by button or drag and drop. - Preview audio with a lightweight waveform cache. - Play, pause, stop, and seek within the waveform. - Add markers and comments to useful moments. - Set a single range per audio file and restore it later. - Textize either the current audio or checked multiple audio files. - Export TXT by default. - Export SRT when needed, including original timeline timecodes for selected ranges. - Schedule textization for later. - Select input and output audio devices. - Export recorded audio files for use in other audio tools. - Switch the interface language between Japanese, English, Chinese, and Korean. IT & Security Facts - Audio is processed locally on your Windows PC. - CowriteAudio does not intentionally upload your audio files. - No telemetry. - No account required. - No admin rights required for normal use. - No auto-update. - Some speech models or dependencies may require a first-time download or local cache setup depending on your environment. Compatibility & Stability - Windows 10 / 11 - x64 - ZIP / portable-style distribution - For 8GB RAM business PCs, the small model is recommended. - Medium and large models may require more memory and longer processing time. - Long audio previews use waveform caching so the interface stays responsive while waveform data is prepared. - Transcription speed depends on your CPU, GPU, RAM, model, output format, audio length, and background apps. - The first transcription may take longer than later runs because the app needs to initialize local settings and load the speech model. This is normal for the first run. Windows SmartScreen / first-run notice Because this is an independently distributed Windows app, Windows SmartScreen may show a first-run warning. If you trust the file, click "More info" and continue. If you are unsure, do not run it and contact support first. 8GB RAM business laptop test CowriteAudio was tested on a regular Windows 11 Pro business laptop with 8GB RAM and an Intel Core i5-1135G7. Test file: - WAV - 42 minutes - 229MB - 48kHz / 16-bit / mono Result with small model: - TXT only: 9 min 55 sec - TXT + SRT: 11 min 30 sec In this test, CowriteAudio processed a 42-minute WAV file into TXT + SRT in about 27% of the audio length, or roughly 3.7x real-time speed. Processing time varies depending on your PC, audio length, model, output format, and background apps. First-time setup or model loading may also take additional time. Support admin@shakado.tools Please include: - App name: CowriteAudio - Windows version - What you were trying to do - Any error message, if available Included in the download The exact ZIP contents may vary slightly by release, but the package is expected to include: - CowriteAudio.exe - _internal/ - ui/ - README_FIRST_EN.txt - README_FIRST_JA.txt - CHANGELOG.txt - LICENSE.txt - shell.json - SHA256.txt SHA256.txt is included so you can verify the downloaded executable.