🎙️ MegaWhisper

Powerful speech-to-text tool for Linux

---

Voice-to-text application for Linux with local Whisper and OpenAI API support.

Features

• Local Whisper — works offline, your data stays on your machine
• OpenAI API — cloud transcription via Whisper API
• Model download — download models directly from the app
• System tray — runs in background with quick access
• Hotkeys — start recording with keyboard shortcuts
• Recording history — all transcriptions are saved
• Translation — built-in translation via GPT

Installation

Download

Download the latest version from Releases.

Run

chmod +x megawhisper
./megawhisper

Application menu shortcut is created automatically on first launch.

System Requirements

• OS: Fedora Linux
• Desktop: KDE Plasma, GNOME (Wayland)

Install dependencies

sudo dnf install qt6-qtbase portaudio flac

Usage

1. Launch the app — icon appears in system tray
2. Open settings (right-click → Settings)
3. Choose mode: Local Whisper or OpenAI API
4. For local mode — download a model
5. Press hotkey (default Ctrl+Shift+Space) and speak
6. Text appears in active window or clipboard

Local Whisper Models

Model	Size	Description
tiny	~75 MB	Fastest
base	~142 MB	Fast
small	~466 MB	Balanced
medium	~1.5 GB	Accurate
large-v3	~3 GB	Latest

For Russian language, medium or large-v3 recommended.

Screenshots

Recording History

Settings