PantryPin v0.1.0

PantryPin v0.1.0 — From photo to pantry

PantryPin is a first public MVP exploring a simple idea: what if a food photo could help you understand likely ingredients, estimate nutrition, compare the cost of recreating the dish, and find useful cooking resources?

Try it

• Live simulated demo
• Source and setup

This is a GitHub source release. GitHub provides downloadable source archives for the tagged version; PantryPin is not published as an npm package.

Highlights

• Food-photo upload flow
• Dish, ingredient, calorie-range, and macro result views
• Confidence labels and visible uncertainty
• ZIP-sensitive simulated basket comparisons
• Links to searches on established cooking sites
• Responsive GitHub Pages deployment
• Optional OpenAI vision backend for self-hosted deployments
• Tests, typed data boundaries, and automated deployment
• MIT-licensed source with contribution and support guides

Know before you try it

The public Pages build uses clearly labeled canned scenarios; it does not inspect image pixels. Store prices, locations, and pickup details are illustrative rather than live retailer data. Self-hosting with an OpenAI API key enables vision-based food estimates, while retailer pricing remains simulated.

A photo cannot reveal exact portions, hidden oils, or every ingredient. Results are estimates and are not medical advice.

Run locally

git clone https://github.com/nahin333/where-and-how-much.git
cd where-and-how-much
npm install
cp .env.example .env.local
npm run dev

Feedback and contributions are welcome. If PantryPin's direction is useful to you, consider starring the repository—or fork it to explore analysis, grocery-data providers, evaluation, or UX.