Nutrition API · private preview
Food recognition API for meal photos, voice and text.
Let your users log a meal the way that suits them. BiteRight recognises the food, estimates the portion and sends back calories, macros, vitamins and minerals as JSON for your app to show its own way.
Private preview. The team confirms fit, sandbox access and setup; an enquiry does not issue keys. Production access, final specifications and pricing are not available yet.
For a patient-review workspace, explore the separate dietitian portal. Confirm portal access and any connection to your own product with the team.
Three ways to log a meal through the API.
Photo and text share one food-log endpoint; voice has its own. Requests are authenticated multipart uploads over HTTPS.
Meal photo recognition
Send one or more meal photos with the date and a portion size. The API identifies the dish, lists likely ingredients and estimates the total weight before looking up nutrients.
images[] · size: Small, Medium, Large or Custom
Voice food logging
Send a voice note such as “two eggs, toast and a flat white”. The voice endpoint turns the description into a food log. It needs a dietary profile and daily intake (RDI) setup first.
voices[] · for example .m4a
Voice food logging API →Text and natural-language logging
Send what the user typed, from a single food to a full meal description. Text works well when a photo is unclear or the user knows exactly what went into the dish.
meal_name · same endpoint as photos
From a meal to a nutrient log.
Recognise
Identify the dish in the photo, voice note or text.
Decompose
Break custom and mixed dishes into likely ingredients instead of relying on an exact dish name.
Estimate the portion
Estimate the total weight, guided by the portion size your app sends.
Retrieve nutrients
Match each food to curated food-composition records and return the nutrients as JSON.
Food data is the foundation. Machine learning interprets the meal; the nutrient values come from curated food-composition records, not from a black-box AI answer. Grounding results in source data reduces the risk of invented nutrient values, though portions and custom-dish composition remain estimates.
Method as described by BiteRight. Nutrient values come from food-composition records; see the food-data sources and how accuracy is assessed.
What the nutrition API returns.
Each meal comes back as structured JSON your app can store, chart or check against its own targets.
- Recognised dish
- The dish the meal was matched to, with the ingredients found in it.
- Nutrient amounts
- Calories, macronutrients, vitamins and minerals, each with an amount and unit, plus dietary information.
- Estimated weight
- Total portion weight, guided by the Small, Medium, Large or Custom size you send.
Energy and macronutrients
Calories, protein, carbohydrates, total fat, saturated, monounsaturated and polyunsaturated fat, sugar, salt, cholesterol and fibre.
Vitamins
Vitamins A, C, D, E and K, and the B vitamins.
Minerals
Calcium, iron, magnesium, potassium, zinc, phosphorus, selenium, copper, manganese, chromium and molybdenum.
Other nutrients and dietary flags
Omega-3, omega-6, caffeine and other listed compounds, plus dietary information such as vegetarian or gluten-free.
Nutrient groups covered by BiteRight’s food data. Which nutrients a food log includes depends on the food and its source records; the sandbox documentation confirms the exact fields, and the specification is not final. Values are estimates.
Where teams use food recognition.
- Hospital and clinic apps
- Evaluate meal logging in your own app. The separate dietitian portal reviews connected BiteRight app records; confirm any connection to your product with the team. Dietitian portal →
- Diabetes and weight programmes
- Your app can compare estimated carbohydrates, sugar and fibre with the targets your programme sets. Your app's targets and alerts →
- Wellness and fitness apps
- Add photo, voice and text logging without building a food database or recognition model. API structure →
- Chronic care programmes
- Put a patient’s meals into context during education and follow-up, with your care team deciding what the estimates mean. Business overview →
- Dietitian software
- Receive meals as structured ingredients and nutrients, so less time goes into cleaning free-text food logs. What comes back →
- Nutrition research
- Collect photo, voice and text food input in one consistent format for later analysis. Validate estimates against your study protocol. How accuracy is assessed →
Potential applications, not customer endorsements. Hosting, encryption and the data processing agreement are covered on the security page.
Food recognition API questions.
Tell us what your users eat.
Share your product, the inputs you need, typical cuisines and expected volume. The team reviews the fit and confirms sandbox access and setup.
Please leave out patient or personal health information.