Can generative AI help us to cook better?

A few weeks ago, Mumbai-based chef and content creator Natasha Gandhi took the help of Google Gemini to come up with recipes using leftover chapatis. In the Instagram video, she asks the AI chatbot—“Gemini, I have leftover chapatis. How can I make a new, tasty and quick snack?” Gemini does a “your wish is my command”, and what follows are recipes for chapati chips, chapati enchiladas and stir-fried chapati noodles. Gandhi gets cracking with her pots and pans. Her followers want to cook them all.

When the team at Hunger Inc., the Mumbai-based hospitality group was brainstorming for Christmas-themed treats, they took to AI tools for ideas. Founder and COO Yash Bhanage used a simple prompt: “A Gingerbread Man inspired by an Indian halwai, fun, playful, mithai-style character.” Within minutes, it generated visuals of a gingerbread man, only in this case, with a moustache and wearing a dhoti.

The time is here when chefs, content creators and food businesses around the world are seeking the help of generative AI to stay ahead of the game. It often serves the push they need to generate new ideas in the form of innovative recipes, menu planning and plating dishes. AI-powered applications can now assist recipe creators by decoding ingredient information and flavour combinations, and suggest unique pairings that they can test and develop. Prompts work like an accelerator and the response for practically everything is out there. While cooking essentially is a human experience, and as a culture much of it relies on instinct, what is the scope for creativity? More importantly, can a machine inspire a human to create a stellar dish?

Gandhi’s feed features a diverse range of recipes, from marination hacks to Chettinad-style Maggi. For content creators like her, being creative and inspired is serious business. “For me, AI is a thought starter. It’s a spark, a framework and a direction,” says Gandhi, who has 1.6 million followers on Instagram. “It can assist, but not inspire. It can give suggestions like, ‘if the salt is more, add acidity or starch’ or ‘if your marinade is too thin, thicken it with nut paste.’ These are helpful, but they’re not inspiration,” she says.

But, AI is a gamechanger when it comes to minimising time and effort. While building a healthier version of a dish, it helps her with protein counts, swaps and portion logic, which earlier used to take time. Gandhi insists the final flavour call is always hers. “Instinct is the base recipe. AI just expands the possibilities. It gives me variations I may not think of immediately, like turning a Maharashtrian thecha into a marinade, or showing me global techniques for an Indian ingredient.”

In another part of the world, chef Grant Achatz of Chicago’s Michelin-starred restaurants Alinea and Next has often found inspiration in ChatGPT for planning menus (reports a New York Times article from June). He has used it to design a nine-course meal for Next to be launched in 2026. He has also relied on the AI tool for ideas ranging from how to use traditional cooking fuels, to paleontology, which even got him to create a dish inspired by Patagonian fossils at Alinea.

While machine-assisted creativity may elevate dining experiences, chefs don’t see AI replacing the instinct or emotion of cooking. “I see it as something that can support operations rather than influence creativity. In theory, it can help project ingredient needs, reduce waste by analysing past data or keep track of large recipe libraries or SOPs, which is useful when you are running multiple kitchens,” says chef Hussain Shahzad, executive chef at Hunger Inc.

Some of his best dishes have been inspired by his travels, from conversations, memories, learnings over the years or even from tasting something by mistake during prep. “These moments carry emotion, memory and context, which a machine doesn’t have access to. So, while AI may nudge me towards a new direction, it cannot replace my intuition and lived experience,” Shahzad says.

Meanwhile, researchers have been working on developing AI-driven recipe generators that will employ large pools of existing recipe data to create new and customised ones as per the user’s requirements. The algorithms are being trained to factor in flavour profiles, cuisines, nutritional value and dietary restrictions.

It was a simple question—can machines imagine new recipes the way chefs do?—that led Ganesh Bagler to build an AI recipe generator called Ratatouille. The Indraprastha Institute of Technology, Delhi professor and computational researcher has been working on what he calls computational gastronomy for over a decade, using data science and algorithms to understand ingredient pairings and flavour combinations. “AI can generate new dishes, personalise nutrition, reduce salt/sugar/oil intelligently, help food brands formulate healthier products, and support sustainable ingredient transitions—all at scale, in minutes,” says Bagler, who along with his team has built databases consisting of 584,000 recipes from over 100 countries, mapping ingredients, nutritional values and carbon footprint.

If a user wants to try a dish that lists Atlantic salmon in the ingredients, can Ratatouille offer substitutes with local catch? “Yes, as long as we provide fine-grained data of carbon footprints of ingredients, it will be able to suggest new recipes as well as modified versions with suitable alternatives,” he says. Bagler stresses that Indian cooking thrives on instinct, memory and inherited technique, and AI cannot replace the experience. “AI doesn’t reduce creativity, it augments it by enabling the user to cook smarter, healthier or more imaginatively.”

Food brands are also leveraging the benefits, especially for product innovation. Ruchira Sonalkar, co-founder of Native Tongue, which retails artisanal condiments and offers B2B contract manufacturing for private-label products, uses artificial intelligence for R&D. “Most clean label brands steer clear of MSG. So how do I get the umami going? If I feed my requirements on ChatGPT, it will give me a precise list of ingredients like mushroom powder and nutritional yeast as they mimic the umami flavour,” she says.

Artificial intelligence allows the Hunger Inc. team to align faster without diluting the originality of the idea. “It works like a rapid visualiser, it translates imagination into something concrete within minutes, which is incredibly powerful in a fast-moving creative environment,” says Bhanage.

The future is human creativity supported by AI efficiency, and not one replacing the other. For now, Bagler and his team are busy collaborating with culinary schools (Institute of Hotel Management, Pusa; Le Cordon Bleu, India; and Symbiosis School of Culinary Arts, Pune) to validate the AI-generated recipes and assess how Ratatouille can help improve their performance in the kitchen. The application will be available for commercial use by mid-2026. Ask him if he’d cook a meal using AI for his family, and he says, “Of course, but I will still do the final tempering of ghee and spices.”

Recently, I prompted ChatGPT—“can I cook butter chicken without butter?” It gave me options like coconut cream/milk. If only it knew my intuition simply won’t accept it.