A privacy-first web app that scans product ingredient lists and tells you if they're vegan or vegetarian. All processing happens locally in the browser — no cloud, no tracking.
🌐 Try it live → macusercom.github.io/v-checker
- Take a photo or upload an image of any ingredient list
- OCR powered by Tesseract.js (German, English, French, Italian, Spanish)
- Checks against a database of 200+ E-numbers and 60+ ingredient keywords
- Classifies as: Vegan / Vegetarian / Not vegetarian / Conditional
- DE/EN interface language switch (independent of OCR language)
- Allergy trace detection — "May contain" sections are flagged as info, not counted for the verdict
- Standalone E-number lookup with bilingual names
- Image preprocessing with adjustable threshold for better OCR results
- No data ever leaves your device
- Open macusercom.github.io/v-checker in your browser
- Tap Take photo or Choose from gallery and select an image of the ingredient list
- Select the language of the text on the packaging (OCR Language)
- Adjust the threshold slider if needed to improve contrast, then tap Analyse →
- Read the verdict and review any flagged ingredients
No install required. Works on mobile and desktop.
Tesseract.js requires a local server — opening index.html directly as file:// will not work.
git clone https://github.com/Macusercom/v-checker.git
cd v-checker
python3 -m http.server
Then open http://localhost:8000 in your browser.
- Vanilla HTML, CSS, JavaScript — no framework, no build step
- Tesseract.js for in-browser OCR
- All ingredient and E-number matching runs locally against a bundled database
No data ever leaves your device. Built as a vibe coding project with Claude.
