https://github.com/mrrfv/workout-buddy-web

Mute the music of your favorite workout video and play your own—without skipping a rep.
https://github.com/mrrfv/workout-buddy-web

ai fitness gemma llm ministral mistral nextjs react workout

Last synced: 16 days ago
JSON representation

Mute the music of your favorite workout video and play your own—without skipping a rep.

• Host: GitHub
• URL: https://github.com/mrrfv/workout-buddy-web
• Owner: mrrfv
• License: agpl-3.0
• Created: 2025-10-24T15:08:51.000Z (8 months ago)
• Default Branch: master
• Last Pushed: 2026-06-10T19:29:41.000Z (24 days ago)
• Last Synced: 2026-06-10T21:10:55.973Z (24 days ago)
• Topics: ai, fitness, gemma, llm, ministral, mistral, nextjs, react, workout
• Language: TypeScript
• Homepage: https://workout-buddy.1998.ovh
• Size: 825 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Workout Buddy

![Demo](./web/public/demo.webp)

Mute the music of your favorite workout video without skipping a rep—play yours instead. Workout Buddy uses on-device vision language models to monitor the video and send instant reminders to switch exercises or take a break. Next.js rewrite of my original, rusty [Workout Buddy](https://github.com/mrrfv/workout-buddy).

## Features

- Video-agnostic. It works with almost any YouTube video, as long as it contains a countdown timer

- Ignores irrelevant scenes (e.g. other tabs, windows)

- Runs entirely in-browser, doesn't permanently store screen captures

    - Works cross-platform

    - You can share only the YouTube tab with the model for better privacy

- Communicates with the AI provider of your choice

    - Ollama

    - OpenRouter

- Works great with the smallest vision language models, like Ministral 3 3B

    - Analyzes 3 screenshots at once, allowing the model to find less obvious timers

- Plays an audible sound with a notification when the countdown is over

## Running locally

The app at `web/` is a standard Next.js project. To run it locally, run `pnpm install` and `pnpm dev` inside the folder.

## Why not OCR?

I realized early on that this problem could've been solved with OCR. However, using a vision language model removes the friction of having to manually point the engine to the timer's location on the screen. A model can automatically find and interpret the timer, regardless of its position or style. This makes the app much more user-friendly.

## License

AGPL-3.0 License. See [LICENSE](./LICENSE) for details.