How a photo becomes a calorie estimate
1. Recognise the dish
The AI model identifies the dish and its likely ingredients: salmon, quinoa, avocado, dressing.
2. Estimate the portion
It estimates the amount of each ingredient from what is visible on the plate.
3. Calculate nutrients
Calories, protein, fat, carbs, fiber, sugar and sodium are calculated for that portion.
4. You decide
You see the estimate and the ingredient list, change what is wrong and save it.
Where photo estimates work best
- Plates where the main foods are visible: grilled meat or fish, rice, vegetables, eggs.
- Common dishes: pasta, salads, sandwiches, bowls, porridge, soups.
- Restaurant and homemade meals you have no barcode for.
Tips for better results
- Shoot from above in good light so every component is visible.
- Add a short note when something is hidden: “cooked in 1 tbsp olive oil”, “no sauce”.
- Check the portion and adjust it if you know the weight.
- For packaged food, scan the barcode or the nutrition label instead of the photo.
- Save meals you eat often to favorites and repeat them without a new estimate.
What AI can’t see
- Oil, butter and sugar absorbed during cooking.
- Ingredients hidden inside a dish or under a sauce.
- Exact weight. A photo has no scale, so portions are an estimate.
Photo and text results are AI estimates. You can edit every number before saving.
Other ways to log in Munoo
Barcode scanner
Nutrition facts from Open Food Facts, a community product database.
Nutrition label scan
Photograph the label and Munoo reads the values per serving or per 100 g.
Quick text or voice
Describe the meal in a few words and get an estimate.
Manual entry and search
Enter numbers you already know or pick a product by name.