Indian health and wellness startup HealthifyMe has launched an innovative AI-powered feature that identifies Indian food from images to simplify calorie tracking, enabling users to monitor their meal intake more effectively.
Given the diversity of Indian cuisine and the common usage of a Thaali plate, which contains various food items in different portions, tracking meals through images can be quite challenging. A robust model must be able to recognize individual dishes and their portion sizes accurately.
HealthifyMe’s new feature, Snap, allows users to take photos of their meals, instantly recognizing the food items included in the shot. Users can also grant the app access to their photo gallery, where Snap will automatically scan food images. This feature lets individuals capture meal photos and log calories at their convenience.
Regarding privacy, HealthifyMe clarifies that images are processed on-device for food detection before being sent to servers for specific dish recognition. The gallery-based model reportedly offers better performance, as it has additional time for food recognition compared to real-time meal photography.
Once Snap detects multiple items in an image, it prompts users to tap on individual dishes to log them for calorie tracking. An adjustable rectangular box appears, allowing users to focus on different food items with ease.
Co-founder and CEO Tushar Vashisht shared that HealthifyMe initially began as a food-tracking app nearly a decade ago, and this remains the primary function for free users. “Engaging in food logging can be cumbersome since it requires manual entry. Our goal with Snap is to enhance user retention and interaction,” he explained.
Vashisht stated that the startup has attempted image-based recognition in the past, but advancements in generative AI models have made creating Snap more feasible. Currently, the feature can recognize around 150,000 Indian food items.
HealthifyMe reports an accuracy rate of 60-70% in food recognition. In cases where the model fails to identify a dish, users will receive alternative suggestions. Human reviewers assist in identifying and correcting any misclassified items, and users can help improve the model by manually labeling incorrectly recognized photos. Vashisht is optimistic that accuracy will surpass 80% within the coming month.
HealthifyMe is not alone in its quest for advanced food recognition technology. Competitors include Samsung's meal planning platform, which plans to launch a similar feature next year, and Snapcalorie, a startup founded by former Google Lens engineer Wade Norris, backed by investors like Y Combinator, Index Ventures, and Accel.
In the coming weeks, HealthifyMe aims to expand food logging options. With Snap’s launch, users will soon be able to send food images via WhatsApp or tag them on X. The company is also developing a voice input feature and enhancing its AI assistant, Ria.
To date, HealthifyMe has secured over $130 million in funding from investors, including LeapFrog Ventures and Khosla Ventures. They offer a basic subscription plan starting at $4.80 (₹399) monthly, providing access to an AI-driven fitness and nutrition assistant, meal planning, and healthy recipes. Pro plans, starting from $48 per month (₹4,000), include features such as an AI-based meal planner, access to nutrition and fitness coaches, and smart scales.
Higher tiers of Pro plans come with added benefits, including multiple doctor consultations, metabolic panel tests, and continuous glucose monitoring (CGM) devices. With over 200,000 subscribers, nearly 40% of whom are on the Pro plan, HealthifyMe is well-positioned for continued growth. In 2020, the company partnered with Indian food delivery service Swiggy to curate healthy meal options and is currently in discussions with various food and grocery services to leverage its innovative technology.
Ranking
