FoodCal AI — Calories & Macros from a Plate Photo

Overview

FoodCal AI turns a single plate photo into a clear, actionable nutrition report: total calories, protein/carbs/fat/fiber, detected foods with portion estimates, allergens, and dietary signals including halal and pork risk. Built for people and teams who want reliable meal tracking without manual search, weighing, or spreadsheets.

The Problem

Most people who start a diet abandon calorie tracking within two weeks because logging is slow and tedious. Existing apps require typing every item and guessing portions. Allergen and halal checks are rarely integrated, so users must switch tools or ask staff. Generic chat AI can guess calories from a photo but does not return consistent structured JSON, confidence scores, or safe disclaimers.

The Solution

FoodCal AI accepts a photo (or URL/batch), segments foods and estimates portions with multimodal vision, then produces a structured JSON report with confidence, macros, warnings, and three next actions. Optional audio/text context improves portion estimates. Results are designed for quick decisions: log the meal, adjust the next portion, or avoid an allergen—not for clinical diagnosis.

Key Capabilities

• Plate photo → calories, protein, carbs, fat, fiber in one structured report
• Detected items with portion estimates and per-item confidence
• Allergen flags and dietary signals (halal / pork risk, vegetarian, vegan, gluten-free)
• Low-confidence retake guidance when the photo is blurry or unclear (under 60% confidence)
• Multilingual report language: English, Persian, Arabic
• REST APIs: analyze, analyze-url, batch, and feedback for corrections
• Pay-as-you-go wallet metering with top-up prompts when balance is empty

Business Benefits

• Faster meal logging than manual entry or generic chat prompts
• Consistent JSON for apps, kiosks, coaching tools, and meal services
• Fewer allergen and dietary mistakes through explicit flags and disclaimers
• Predictable unit economics on a prepaid Sensifai wallet

Use Cases

• Ideal for consumers tracking daily calories and macros from phone photos
• Suitable for nutrition coaches reviewing client meal photos at scale
• Designed for workplace cafeterias or meal-kit apps that need quick plate estimates
• Can be used by food-tech builders who need a reliable photo-to-nutrition API

Who It's For

Health and fitness consumers, diet-conscious professionals, personal trainers and coaches, and product teams building consumer or B2B food and wellness experiences.

How It Works

1. Capture or upload a plate photo. 2) Vision models identify foods and portion cues. 3) Reasoning models compute calories and macros and apply safety filters. 4) You receive a report card with confidence badge, actions, and optional share/save. Images are intended for short-lived processing (24h retention policy) unless a report is saved.

Why This Application

FoodCal AI is purpose-built for photo-to-nutrition workflows: structured outputs, confidence and retake guardrails, allergen and halal signals, and metered pricing that matches real usage—unlike generic assistants that require custom prompts and still return inconsistent formats.