Camera-first experience
Designed for how you actually eat
Mobile is the primary experience. Every detail of the scanner is built for speed and clarity.
Live framing guide
An on-screen guide helps you capture the full plate for better recognition.
Flash control
Toggle flash for low-light situations. Your photos come through clearly.
Gallery & retake
Upload from gallery or retake instantly. No pressure to get it perfect.
Progressive analysis
Watch each stage of analysis happen in real-time with elegant progress animation.
Confirm or edit
AI results are always editable. Correct food, portion, and preparation.
Manual entry
Don't want to take a photo? Log meals manually with full nutrition data.
Analyzing your plate
What happens after you capture
PlateBeacon processes your meal through four progressive stages.
Identifying foods
The AI scans the image to detect individual food items on the plate.
Estimating portions
Portion sizes are estimated based on visual cues and reference data.
Calculating nutrition
Nutrition is estimated from the identified foods and portions.
Preparing your insights
Personalized insights and suggestions are generated for your meal.
AI architecture
Provider-agnostic by design
PlateBeacon does not hardcode one AI provider. Our architecture supports future integration with multiple vision models.
Multiple providers
Architecture supports OpenAI vision, Gemini vision, specialized food APIs, and custom models.
You stay in control
AI never silently changes user-confirmed food records. Everything is editable.
Estimates, not guarantees
Food recognition is an approximation. We never present it as laboratory-level accuracy.
Plan-based controls
AI usage is managed server-side based on your subscription plan and fair-use limits.