01
Camera photo
Photograph the meal when the plate is in front of you and you want a visual estimate.
AI food scanner
Turn a meal into a useful nutrition estimate
Choose the input that fits the moment: take a photo, select one from your gallery, describe the meal in text, speak it aloud, or scan a barcode.
Available on iOS and Android. Apple Health, widgets, and Siri features are iOS-only.
AI food scanner facts
- Meal inputs
- Photo, gallery, text, voice, or barcode
- Nutrition output
- Calories, P/F/C, fiber, added sugar, and saturated fat
- AI analyses
- Directional Health Index
- Personalization
- Optional editable Food Memory hints
Five input methods
Use the quickest route from plate to log
Meals do not arrive in one neat format. Food Focus gives you five ways to start an analysis so you can choose based on what is in front of you.
02
Gallery image
Use an existing meal photo when you are logging later rather than at the table.
03
Text description
Describe ingredients and portions in words when those details are clearer than the image.
04
Voice description
Speak the meal instead of typing when you need a hands-free starting point.
05
Barcode scan
Scan a packaged product barcode when a product record is the most direct source.
One readable result
See the numbers people use most often
Every flow can show the seven tracked nutrition fields. Photo, gallery, text, and voice AI analyses also produce a directional Health Index; the barcode flow currently initializes that score at zero rather than generating the same assessment.
01
Calories and macros
Review estimated calories plus protein, fat, and carbohydrates for the meal.
02
Fiber and sugars
See estimated fiber and added sugar alongside the main macro breakdown.
03
Saturated fat
Keep estimated saturated fat visible without opening a separate nutrition workflow.
04
Health Index
For AI meal analyses, use the Health Index as directional context, not a clinical score. Do not interpret the barcode flow's default zero as a meal-quality assessment.
Food Memory
Make familiar meals easier to recognize over time
People often repeat the same breakfast, coffee, bowl, or recipe. When enabled, Food Memory keeps editable gram hints for normalized ingredients and can use them as context in future analysis.
01
Portion context
In the normal single-meal flow, Food Memory replaces the current grams for a recognized ingredient rather than averaging or frequency-weighting past values.
02
Useful for routines
The feature is especially relevant when your usual ingredients and serving sizes repeat across days.
03
Still an estimate
Remembered context can support analysis, but it does not turn an image or description into a laboratory measurement.
How to use it
A simple three-step flow
- Step 1
Choose an input
Take or select a photo, type or speak a description, or scan a packaged-food barcode.
- Step 2
Review the estimate
Check calories, macros, fiber, added sugar, and saturated fat. Treat Health Index as an AI-analysis signal, and compare barcode fields with the package.
- Step 3
Build useful context
When Food Memory is enabled, a standard reviewed meal save can update ingredient gram hints for later ambiguous analyses.
Use in context
Important limitations
- Photo, gallery, text, and voice results are estimates. Hidden ingredients, preparation methods, and uncertain portion sizes can change the real nutrition values.
- A barcode identifies a packaged product record. Missing tracked fields may be AI-estimated per 100 g, and the gram amount still needs review.
- Health Index is directional and non-clinical. The barcode flow currently uses a default score of zero rather than the same AI meal-quality analysis.
Evidence-led guide
Which food-logging method fits the meal?
Compare photo, gallery, barcode, voice, and text by the evidence each method can capture and the checks each one still needs.
AI food scanner FAQ
Common questions, clearly answered
What can I use to scan or log food in Food Focus?
You can take a photo, choose a gallery image, enter a text description, speak a voice description, or scan a packaged-food barcode.
What nutrition information does the scanner return?
All five flows can return calories, protein, fat, carbohydrates, fiber, added sugar, and saturated fat. Photo, gallery, text, and voice AI analyses also generate a Health Index. Barcode scans currently use a default zero rather than the same meaningful Health Index calculation.
Does the AI food scanner measure my meal exactly?
No. The result is an estimate based on the available input. Portions, hidden ingredients, and cooking methods can affect the actual values.
What does Food Memory remember?
When enabled, Food Memory keeps editable gram hints for normalized ingredients and may use them when a later photo, text, or voice amount is ambiguous.
Is the Health Index medical advice?
No. When generated for a photo, gallery, text, or voice analysis, the Health Index is an informational heuristic rather than medical advice or a diagnosis. A barcode's default zero is not a meal-quality assessment.
Keep exploring
Related Food Focus features
Reuse current gram hints
Food Memory
When enabled, the normal single-meal save flow keeps a current gram value for each ingredient it recognizes and normalizes. A later ambiguous photo, text, or voice log may use that value as context.
Explore feature →See patterns beyond one meal
Nutrition tracker
Move from a single meal estimate to a clearer view of calories, macros, and selected nutrition signals across today and longer chart ranges.
Explore feature →Get guidance from today's context
AI nutrition coach
After you log food, Food Focus can automatically turn your profile, today's meals, targets, and available activity into a short recommendation.
Explore feature →Try ai food scanner
Bring more context to your next meal log
Download Food Focus for iOS or Android. Nutrition estimates and AI guidance are informational, not medical advice.