Installation

Audiotext is a desktop app for Windows, macOS and Linux. It transcribes the audio of files, YouTube videos and microphone recordings into text, and can translate, summarize and subtitle it.

Download the app

Download the file for your system from the latest release on GitHub:

System	File
Windows (64-bit)	Audiotext-<version>-windows-x64-setup.exe
macOS 15 or later (Apple Silicon)	Audiotext-<version>-macos-arm64.dmg
Linux (x86_64)	Audiotext-<version>-linux-x86_64.tar.gz

The app includes everything it needs, FFmpeg included.

Windows

Run the installer and follow its steps. It doesn’t need administrator permissions. If you have an NVIDIA GPU, check Use the NVIDIA GPU (CUDA) to transcribe faster: the installer then downloads the GPU add-on, which makes WhisperX much faster. The installer isn’t signed, so Windows SmartScreen may warn about it: click More info → Run anyway.

macOS

Open the .dmg file and drag Audiotext to the Applications folder. The app isn’t notarized by Apple, so macOS blocks it the first time you open it: go to System Settings → Privacy & Security and click Open Anyway next to the message about Audiotext. On Macs, WhisperX transcribes on the GPU of Apple Silicon with MLX Whisper, much faster than on the CPU, without any add-on. Intel Macs aren’t supported, since PyTorch no longer supports them.

Linux

Extract the archive and run the installer from a terminal:

tar -xzf Audiotext-<version>-linux-x86_64.tar.gz
cd Audiotext-<version>-linux-x86_64
./install.sh

It installs Audiotext for your user and adds it to the applications menu (it can also be opened with the audiotext command). If it detects an NVIDIA GPU, it offers to download the GPU add-on. Run ./install.sh --gpu or ./install.sh --cpu to choose without being asked, and ./install.sh --uninstall to uninstall it (your settings are kept).

Tip

The GPU add-on is a download of about 2 GB on Windows and 4 GB on Linux, so it’s only worth it with an NVIDIA GPU. Without it, WhisperX runs on the CPU, and the Whisper API and the Google API work the same. To switch between the CPU and the GPU versions later, install the app again and choose the other option.

Note

The first time you transcribe with WhisperX (the default engine), its model is downloaded. It takes from ~75 MB for tiny to ~3 GB for large-v2, so it may take a while. The next transcriptions start right away.

Requirements

• WhisperX runs on your computer. It works on any CPU, but it’s much faster on an NVIDIA GPU with CUDA or on the GPU of a Mac with Apple Silicon. See Engines to choose a model that fits your hardware.
• The Whisper API and the Google API run on remote servers, so they need an Internet connection but no powerful hardware.
• To transcribe from the microphone, your system must detect an input device.
• On Linux, recording and playing audio need PortAudio (sudo apt install libportaudio2 on Ubuntu or Debian).

Update the app

When a new version is released, Audiotext shows a Version … is available button in the top bar. Click it to open the page of the release, download the file for your system and install it like the first time: run the new installer on Windows, drag the new app to the Applications folder on macOS, or run the install.sh of the new archive on Linux. The previous version is replaced, and your settings and history are kept, since they’re stored in your user configuration folder. If you use the GPU add-on, choose it again when installing.

To check for a new version yourself, open Preferences → About → Check for updates. To stop checking when the app opens, turn off General → Check for updates automatically.

Uninstall the app

• Windows: uninstall it from Settings → Apps, like any other app.
• macOS: drag Audiotext from the Applications folder to the Trash.
• Linux: run ./install.sh --uninstall from the folder of the archive. If you no longer have it, delete ~/.local/share/audiotext, ~/.local/bin/audiotext and ~/.local/share/applications/audiotext.desktop.

Your settings, history and recordings are kept, so they’re still there if you install it again. To remove everything:

1. Before uninstalling, remove your API keys in Preferences → API keys (click Change… and leave it empty), since they’re kept in the credential store of your system.
2. Delete your user configuration folder.
3. Delete the models that were downloaded.

Change the language of the interface

Audiotext uses the language of your system if it’s available. To change it, open the Preferences (the gear at the top right) and choose a language in General → Interface language.

Run it from the source code

If you want to run the latest code or contribute, see Contributing to set up the project with Python.

Next steps

• Your first transcription explains the window and the steps to transcribe.