Termux Whisper 🗣️

Private, Offline Audio Transcription for Android

A lightweight wrapper for whisper.cpp to transcribe audio files locally on Android via Termux. No internet required.

🚀 Quick Start

1. Install & Setup (One-Liner)

Open Termux and paste this command:

curl -sL https://raw.githubusercontent.com/itsmuaaz/termux-whisper/main/install.sh | bash

This will install all dependencies, fetch the engine, and compile it automatically.

---

2. Manual Installation (Alternative)

If you prefer to do it step-by-step:

pkg install git -y
git clone https://github.com/itsmuaaz/termux-whisper.git
cd termux-whisper
chmod +x core/*.sh menu.sh
./core/setup.sh

3. Usage

Interactive Dashboard: Restart Termux, then simply type:

whisper

CLI Commands: You can also use the whisper command directly with arguments:

Goal	Command
Transcribe File	whisper video.mp4
Specify Model	whisper voice.m4a --model base
Generate Subtitles	whisper movie.mkv --subs
Generate Lyrics	whisper song.mp3 --lrc
Disable Text Output	whisper video.mp4 --no-txt
Disable Paragraphs	whisper video.mp4 --no-paragraphs
Batch Process	whisper /path/to/folder/
Direct File Picker	whisper --file-picker

Available Models: tiny, base, small, medium, large-v3-turbo

4. Interactive Pre-Flight

When you run Whisper (via menu or Share), you will see a 20-second "Pre-Flight" dashboard.

• Shows your current target, model, and output settings.
• Allows you to toggle Subtitles/Lyrics or change the model for this run only.
• Press Enter to start immediately.
• Note: Using specific CLI flags (e.g., --subs) skips this check for faster automation.

✨ Features

• One-Click Dashboard: A professional TUI menu to access all features.
• Share to Transcribe: Automatically enabled. Share audio files from WhatsApp or File Manager directly to Termux.
• Privacy: 100% offline; data stays on your device.
• Smart: Auto-detects audio formats (including OPUS/OGG) and checks for audio streams.
• Native Picker: Visual file picker via Android System (--file-picker).
• Live Progress: Real-time feedback showing transcription segments and timestamps.
• Smart Paragraphs: Automatically structures text into paragraphs for better readability.
• Convenient: Prompts to open the transcript immediately after processing.
• Batch: Transcribe single files or entire directories.
• Subtitles: Optionally generate .srt and .vtt files.
• Lyrics: Generate .lrc files for synced music lyrics.
• Formats: Supports MP3, WAV, M4A, OPUS, OGG, FLAC, MP4, MKV, AVI, MOV.

⚠️ Notes

• Transcripts: Saved to /sdcard/Download/Termux-Whisper/ when using the picker, or next to the file if run manually.

🌟 See Also

Check out Termux Bootstrap – A modular, safe, and mobile-optimized script to transform Termux into a powerful development environment. It includes this project as a community extra!

---

Powered by whisper.cpp