Turn food photos
into nutrition data.
Instantly.
CaloAI converts any food photograph into structured nutritional JSON in under 400ms. Skip months of model training and integrate visual food intelligence in a single API call.
{
"product_name": "Grilled Chicken Rice Bowl",
"calories": 506,
"proteins": 38,
"carbs": 48,
"fats": 18,
"healthy_points": 8,
"confidence_score": 96,
"detected_dishes": [
{
"name": "Grilled Chicken Breast",
"weight_g": 150,
"calories": 156
},
{
"name": "Brown Rice",
"weight_g": 120,
"calories": 186
},
{
"name": "Steamed Broccoli with Olive Oil",
"weight_g": 80,
"calories": 164
}
],
"diet_tags": ["high-protein", "clean-eating"],
"potential_allergens": ["none"],
"ai_verdict": "An exceptional post-workout meal."
}Three meals.
Three real API responses.
Every field shown here is the actual JSON our API returns. Nothing is fabricated for the demo.
detected_dishes[0]
Teriyaki Chicken Bowl
ai_verdict
“Lean glazed protein, complex carbs from rice, and micronutrient-rich broccoli. Caloric distribution favors sustained energy with a low fat load.”
Powering teams at
Three steps from photo to production feature.
Integrate in an afternoon. Your first result in under 10 minutes.
Send a food photo
POST a base64 image to the REST endpoint. Include the extra field with text context for ground-truth portion overrides.
VLM analyzes and validates
Our Vision Language Model estimates portions and ingredients, then validates every calorie against the thermodynamic consistency formula.
Receive structured JSON
A typed JSON object arrives in under 400ms, ready to plug directly into your app state. No parsing, no cleanup, no surprises.
Proprietary tech
built for food.
CaloAI runs on a custom VLM trained exclusively on 150,000+ global dishes, ingredients, and biochemical nutrient profiles. Not a wrapper around a general-purpose vision model.
Validation formula
Cal = (P×4) + (C×4) + (F×9)Thermodynamic Validation
Every response is checked against Calories = (P×4) + (C×4) + (F×9) before it leaves the API. The numbers always add up correctly.
Structured JSON Output
Native token serialization forces typed output. No markdown wrappers, no fragile regex parsing. Structured data, ready to render directly.
Context Fusion
Pass text metadata in the extra field ("Pumpkin soup, 250g") and the model treats it as ground truth, running exact micro-nutrient calculations from the text.
Start free. Unlock more
data as you grow.
Every tier runs the same VLM. Higher plans unlock more response fields, not better accuracy.
Test the API and prototype your MVP
Start FreeShip production nutrition features from day one
Get StartedScale alongside your growing user base
Get StartedEnterprise performance, zero concessions
Get StartedAll plans include Thermodynamic Validation · Enterprise JSON · Context Fusion · No credit card for BASIC
Built for every platform
that touches food.
Fitness apps
Users log meals by photograph in under a second. Macro breakdowns and diet tags arrive instantly, without manual entry.
Health platforms
Allergen alerts and nutritionist verdicts differentiate your premium tier. Build clinical food journals without training a vision model.
Corporate wellness
Food intelligence for employee health programs. Measure nutrition patterns across a team and surface data-backed recommendations.
Your users deserve
instant food intelligence.
CaloAI gives you the JSON. You write the feature. First API key in 60 seconds, free.
Upgrade to PRO anytime · Cancel anytime · GDPR compliant