Instacart has joined the delivery-app AI race with Clementine, a conversational assistant that builds a ready-to-buy cart from plain-language requests. Shoppers can ask for a week of budget-friendly kids’ lunches, high-protein dinners for two, or simply order my usuals, and the assistant assembles the order.
Clementine draws on roughly fifteen years of Instacart purchase data to generate recipes or surface them from trusted publishers, then filters recommendations by household preferences such as gluten-free, vegetarian, nut-free or organic. It also flags deals and promotions and suggests cheaper alternatives when a budget matters.
Inputs go beyond typed prompts. Users can photograph a handwritten grocery list or screenshot a digital one, and the assistant converts the image into a cart. Chief executive Chris Rogers said Clementine puts the company’s knowledge of how families shop and eat to work on the universal question of what to make for dinner.
The launch is available in the US and Canada, and it makes Instacart the latest platform trying to become the household planning layer rather than just a delivery pipe. Uber Eats and DoorDash already offer similar assistants, and Shipt shipped its own version, Ask Shipt, the same day. The common goal is keeping meal decisions inside each app instead of losing shoppers to general-purpose chatbots like ChatGPT.
Whether shoppers trust an assistant to fill their carts, rather than merely suggest items, is the open question as these platforms compete to own the moment before a purchase.
