Calorie Counter

Every calorie counter fails the same way: too many taps to log a meal. This one was designed around the idea that logging food should take less than five seconds, no matter how you want to do it. Scan a barcode, snap a photo, say what you ate, or type it. The OpenAI API handles the recognition and returns calories plus full macronutrient breakdowns for fat, carbs, and protein.

Barcode scanning uses the device camera to read product codes instantly, pulling nutrition data from OpenFoodFacts and filling in the gaps with AI when needed. Photo input sends an image of your plate to the OpenAI Vision API, which identifies the food items and estimates portion sizes. Voice input lets you describe your meal in natural language and get a parsed entry back in seconds.

The tracking interface is built around a tabbed view that lets you switch between calories, fat, carbs, and protein. Each macro has its own daily goal, progress bar, and trend chart. The history view shows interactive graphs over time so you can spot patterns in your eating habits without digging through spreadsheets.

All data is stored locally on your device using IndexedDB. Nothing goes to a server. The app works offline as a Progressive Web App and can be installed on any device. When you want your data out, CSV export gives you everything in a format that opens in any spreadsheet tool.

The interface is deliberately minimal. Dark theme, clean typography, large touch targets, and zero clutter. There are no ads, no premium upsells, no social features. It does one thing and does it fast.