Instacart launches Clementine, an AI assistant that turns recipes and lists into carts

Instacart launches Clementine

Instacart has introduced Clementine, a conversational AI assistant that turns a chat, a recipe or a shopping list into a ready-to-buy cart. According to TechCrunch, the company announced the feature on Wednesday and has made it available to users in the United States and Canada.

The assistant is aimed at the decisions that happen before checkout: what to cook, what to buy and how to keep spending in check. TechCrunch points to example prompts such as asking for a week of inexpensive children's lunches, high-protein dinners that are easy to make for two, or a simple reorder of a household's usual items.

From recipes and lists to carts

Clementine can either generate recipes of its own or surface them from publishers Instacart trusts, TechCrunch reports. Its recommendations adapt to each household's preferences, covering requirements such as gluten-free, vegetarian, nut-free or organic.

The assistant also handles input that is not typed into a chat. Users can photograph a handwritten grocery list or screenshot a digital one, and Clementine will convert it into a cart. As it builds a cart, it highlights relevant deals and promotions and proposes lower-cost alternatives, which makes budget control a core part of the product rather than an afterthought.

Instacart chief executive Chris Rogers described the problem the assistant is meant to solve in the launch announcement, as quoted by TechCrunch: every night, millions of families face the same question of what to have for dinner, and often have no answer, let alone a shopping plan. He added that the company has spent nearly 15 years learning how families shop and eat, and that Clementine puts that knowledge to work.

Delivery apps want the planning stage

Instacart is not the first delivery platform to move in this direction. TechCrunch notes that Uber Eats and DoorDash have both introduced AI assistants capable of converting natural-language requests and grocery lists into personalised carts.

Taken together, these launches signal a change in what the apps are for. They are no longer competing purely on delivery speed and catalogue breadth; they want to be the default tool households open when planning meals and groceries. There is also a defensive dimension: by building planning features in-house, the platforms keep users away from general-purpose chatbots such as ChatGPT, which have become a natural place to ask what to cook tonight.

Why it matters

Clementine is a bid to capture the earliest stage of the shopping funnel. If a household decides what to eat inside Instacart, the resulting cart and the order likely follow. An assistant that suggests recipes, applies promotions and proposes cheaper substitutes shortens the distance between vague intent and a paid transaction, and each step it absorbs is a step users no longer spend in a search engine, a recipe site or a rival app.

The competitive stakes go beyond a single feature. Uber Eats and DoorDash are making the same play, and Instacart is betting that its years of household shopping data, the asset its CEO points to, will make its assistant more useful than a generic chatbot. Whichever tool becomes the habitual answer to the nightly dinner question gains a position at the very start of household grocery spending.