Each macro against your daily target, not a generic reference. One plate carried half the day's protein and an eighth of its fat.
Every app can count calories.
Ours can read the meal.
Photo, barcode, voice or text. One log returns energy, digestion and glycemic impact, with every value linked to its source.
Download the app639 kcal. Never just calories.
One logged chicken salad, read three ways: what it was worth in energy, what it means for a sensitive gut, and what its glycemic load actually was.
Understand your meals6 ingredients checked
- chicken
- couscous
- mixed greens
- cherry tomatoes
- cucumber
- red onion
A verdict on the plate you logged, not on the recipe's name. Red onion and couscous both carry fructans; at this portion they read moderate rather than high.
The couscous carried it. An index of 68 is middling; a load of 46 is not, and the difference between them is the portion. None of it is sugar.
Three readings from one log. And the log is one photo.
Logging without the ritual
In, in seconds.
Read in full.
Four ways to log. Every one comes back as the same complete reading: calories, macros and micros, FODMAP context, glycemic load.
- Photo One frame, whole plate. Portions are marked EST and editable to the gram, and corrections recalculate everything.
- Label shot Box diet or a packaged meal with a printed label? Photograph the label. We read it and compute the ingredients. No retyping.
- Barcode Packaged products with verified, source-marked values.
- Voice “Just had a roll with a banana.” Said out loud, then logged, computed, read.
The app itself
One log in.
Three ways to read it.
Real screens from a real logged meal, not a concept.
- 01What you logged
A photograph, the calories, and the three macros against the day you are having. 639 kcal, and 63 g of protein out of 128.
- 02Down to which part of the plate
Protein attributed to every detected ingredient, in grams and as a share. The chicken carried 46.5 g of the 62.8 — very nearly three quarters of it.
- 03And what it adds up to
1,718 kcal a day against a 2,200 target, with the macro averages underneath. Fat is over; protein is not.
Protein, with provenance
Every app says 63 g.
Ours says where it came from.
A total is a number you cannot act on. The share is. In this chicken salad the chicken carried three quarters of the protein and the couscous most of the rest, while the greens, tomatoes, cucumber and onion came to four per cent between them.
| Detected ingredient | Protein | Share of total |
|---|---|---|
| Chicken | 46.5 g | 74% |
| Couscous | 14.3 g | 23% |
| Mixed greens | 1 g | 2% |
| Cherry tomatoes | 0.4 g | 1% |
| Cucumber and red onion, together from the total | 0.6 g | 1% |
| Total | 62.8 g | 101% |
A real logged meal, not a model: the app detects the ingredients and attributes the protein to each one. The card rounds the meal to 63 g and the breakdown totals 62.8; the shares here are computed from those grams, and sum to 101 after rounding rather than being quietly corrected. Change the portion and the shares move with it, the same way the glycemic load does.
Portion, not the food’s name
Move the serving.
Watch what follows.
Glycemic index is measured on a fixed 50 g carbohydrate test dose. That is what makes foods comparable, and what makes it a poor description of dinner. Drag the slider: the index holds still while everything you actually ate moves with it.
Boiled white rice. Composition from USDA FoodData Central, index from the international GI tables, load computed as index × available carbohydrate ÷ 100.
- Available carb
- 42g
- Energy
- 195kcal
- Protein
- 4.0g
Every value here is published or arithmetic on a published value, never an estimate. Try it on another food.
The trust layer
Know where the
number came from.
A photo can’t see hidden oils or exact grams, and no honest app pretends it can. Four evidence states, printed next to every value. Unknowns say unknown.
- PUBPublished
- a named source: USDA FoodData Central, Univ. of Sydney GI, lab-tested FODMAP data
- CALCCalculated
- GI × available carb ÷ 100, at your portion, not a 100 g average
- ESTEstimated
- the AI’s guess, always editable, and everything downstream recalculates
- —Unknown
- we say so, rather than invent a value
Use cases
Start from what
you’re managing.
The same diary, read through your goal. Four paths, each one about food: no wearables, no workouts, no lab results.
- IBS & gut FODMAP context at the serving you actually ate, not a verdict on the food's name. →
- Weight loss Calories that stay honest: labeled, editable, no invented numbers. →
- Protein & muscle Not just 63 g at lunch, but which part of the plate delivered it. →
- Metabolic health Published index, and a load computed for the grams you logged. →
3sense Food
Get it on your phone
Point your camera at the code. Logging is a photo of the plate, so the app belongs on the device that has the camera.
food.3sense.ai