Calorie Mama
Food AI, by the makers of Calorie Mama
An API that turns a food photo into nutrition facts.
Send it a picture of a meal. It identifies the dishes and returns calories, protein, carbs, fat and more per serving, in a split second. Developers call it directly; the Calorie Mama app is built on it.
Try it with your own photo
Photos are sent to Food AI and may be kept to improve recognition. How we handle them
Raw JSON response
Or try a sample photo
What comes back
The likely dishes with a confidence score, standard serving sizes, and calories, protein, fat and carbohydrates for each. One JSON response, usually within two seconds.
Where it is used
The Calorie Mama app logs meals from photos with it. Teams use the same API for nutrition logging, food databases and restaurant tools.
Get a key
Keys, plans and usage live on the developer portal. If you would rather talk first, the contact form reaches the team at Azumio.
Why not just ask an LLM?
Predictable
Food AI is a dedicated food recognition model, not a chat model. The same photo gets the same answer, every time, in the same JSON shape.
No invented numbers
It picks dishes from a fixed food database and returns that database's nutrition values. It can misjudge a photo, but it cannot make up a food or a calorie count.
Fast and cheap to run
An answer comes back in a split second, at a fraction of the cost of a large model call. That matters when every meal a user logs is a request.