Snap a photo of your food and instantly get an AI-powered carb count, helping people with diabetes track carbs and manage blood sugar with confidence.
CarbScan is an AI-powered carb counting app aimed at people managing diabetes and blood sugar. Users snap a photo of their meal and the app returns an estimated carb count, removing the friction of manual logging or guesswork at restaurants and home.
The product is positioned squarely at the diabetes community: counting carbohydrates is one of the most repetitive and error-prone parts of daily glucose management, and a fast camera-first workflow is meaningfully better than typing food entries into a tracker. CarbScan markets itself as a way to "eat with confidence" by turning carb estimation into a single tap.
The web app is built as a Vite/React single-page application with an installable PWA manifest, so it can be added to a phone home screen and used like a native app. The HTML still contains the replit-dev-banner.js script, which is the fingerprint Replit injects into projects vibe-coded on its platform — strong evidence the app was built with Replit's AI app builder rather than Lovable, as some directories list. Branding uses an Inter typeface and a custom blue droplet icon evoking blood glucose monitoring.
CarbScan is a clear example of a single-purpose, vertical AI utility shipped quickly by a solo or small team using a vibe-coding tool: take an existing model capability (multimodal food recognition), wrap it in a focused UX, and aim it at a community with a real, daily pain point.
Tools
Languages
AI Models
This listing is unclaimed. Sign in to claim it.