Bleep That Sh*t! — In-Browser Audio & Video Censorship Tool

Make someone sound naughty 😈 or make your content more Ad-friendly.

Bleep That Sh*t! lets you instantly transcribe and censor words in your audio or video files, with full control over what gets bleeped. No uploads, no servers, no installs — everything happens right in your browser.

The current in-browser implementation works for clips of length 10 minutes or less.

How it works

1. Upload your audio (MP3) or video (MP4) file.
2. Transcribe using Whisper ONNX models (transformers.js) — all in-browser.
3. Censor: Pick words to bleep (exact, partial, or fuzzy match).
4. Preview & Download: Hear the result and save your censored file.

All processing is done locally in your browser. Your media and transcripts stay private — nothing is sent to a server.

Examples

Some examples of the end product (make sure to turn volume on, its off by default).

https://github.com/user-attachments/assets/da50f8a9-27ba-4747-92e0-72a25e65175c

Let's look more closely at the last example above - below is a short clip we'll bleep out some words from using the pipeline in this repo. (make sure to turn on audio - its off by default)

https://github.com/neonwatty/bleep-that-shit/assets/16326421/fb8568fe-aba0-49e2-a563-642d658c0651

Now the same clip with the words - "treetz", "ice", "cream", "chocolate", "syrup", and "cookie" - bleeped out

https://github.com/neonwatty/bleep-that-shit/assets/16326421/63ebd7a0-46f6-4efd-80ea-20512ff427c0

Using the App

Just open the app in any modern browser — no installation or setup required.

Live Demo

Visit the deployed app at https://neonwatty.github.io/bleep-that-shit/ or run it locally.

App Features

Main Workflow (/bleep)

• Upload audio or video files
• Select language and Whisper model
• Transcribe to generate word-level timestamps
• Enter words to censor with multiple matching modes (exact, partial, fuzzy)
• Choose from different bleep sounds
• Preview and download the censored result

Model Comparison (/sampler)

• Compare transcription accuracy across different Whisper models
• Test processing speeds
• Find the best model for your needs

Running Locally

Prerequisites

• Node.js 18+
• npm or yarn

Installation

1. Clone the repository:

   git clone https://github.com/neonwatty/bleep-that-shit.git
   cd bleep-that-shit
   

2. Install dependencies:

   npm install
   

3. Start the development server:

   npm run dev
   

4. Open http://localhost:3000 in your browser

Available Scripts

npm run dev          # Start development server
npm run build        # Build for production
npm run start        # Start production server
npm test            # Run Playwright tests
npm run test:ui     # Run tests with UI

Browser Requirements

• Modern browser with WebAssembly support
• Web Workers support
• Sufficient RAM for model loading (varies by model size)

Available Whisper Models

• Tiny (39 MB): Fastest, good for quick processing
• Base (74 MB): Better accuracy, still fast
• Small (242 MB): Best accuracy, slower processing

Both English-only and multilingual variants are available.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Acknowledgments

• Transformers.js for WebAssembly Whisper models
• FFmpeg.wasm for media processing
• OpenAI Whisper for the original models

Support

For issues and questions, please open an issue on GitHub.