Amazon is expanding Rufus from a limited beta to a shopping tool available to all U.S. customers in the Amazon mobile app. The AI-powered assistant is built to help shoppers ask questions, compare products, get recommendations and navigate parts of their Amazon account from a chat interface.
The rollout moves Rufus into a more visible role inside one of Amazon's most important customer surfaces: the mobile shopping app. For shoppers, the change means product research can now happen through questions and follow-ups instead of only through search boxes, filters and product pages.
What Rufus does inside the Amazon app
Rufus appears in the bottom right of the Amazon mobile app's main navigation bar. U.S. customers need the latest version of the Amazon Shopping app to access it. The icon uses chat bubbles with a sparkle, making the assistant a persistent part of the app experience rather than a separate destination.
Amazon says Rufus is designed for several shopping tasks. It can help customers find products, perform product comparisons and receive recommendations on what to buy. The assistant can also answer questions that sit around a purchase decision, such as what factors matter before choosing a product or how one item differs from another.
The chatbot was first announced in February and was initially available in beta to select U.S. customers. Amazon says the broader release follows testing across “tens of millions of questions.” That beta period gave the company a large set of shopping interactions to observe before opening Rufus to all U.S. shoppers.
How Rufus answers shopping questions
Rufus is powered by an internal large language model specialized for shopping. According to Amazon, it has been trained on Amazon's product catalog, customer reviews, community Q&As and other public information from around the web. Amazon has not disclosed which websites were used, or whether data from other retail websites was included.
That mix of sources allows Rufus to respond to broader buying questions as well as product-specific ones. A shopper can ask what to consider when buying headphones, what to consider when detailing a car at home, what clean beauty products are, or what is needed for cold weather golf.
The assistant can also respond when a customer describes a goal rather than naming a product. If someone says, “I want to start an indoor garden,” Rufus can suggest items that would help with that task. This makes the interaction closer to a guided shopping conversation than a traditional keyword search.
Amazon says Rufus can draw on customer reviews and other expert analysis found around the web to discuss how well a product holds up. It can also answer questions about product materials and then guide users toward related questions inside the same chat window.
Why guided questions matter
During testing, Amazon found that customers asked Rufus direct questions and also clicked related questions that appeared in the chat window. In one example, a customer might ask, “What's the material of the backpack?,” then tap “What do customers say?” to continue learning about the product.
That follow-up behavior matters because shopping decisions often involve several small checks rather than one simple answer. A customer may want to know what an item is made of, how it compares with another product, what reviewers mention and whether it fits a specific use case. Rufus is built to keep that process in one conversational flow.
The assistant can also bring in context beyond a product listing. Amazon says that when a customer asked about a pool umbrella for Florida, Rufus shared facts about Florida's weather, humidity and more. In that case, the assistant treated the location and use case as relevant to the buying decision.
Rufus is not limited to static product details. Amazon says customers used it to keep up with fashion trends and the latest tech, including asking about the latest model of a product or what styles were popular. For shoppers who do not already know the category well, that can make the first stage of research less dependent on browsing many separate pages.
Order help and known limits
Rufus also served beta customers as an assistant for account-related shopping tasks. It could help users find past orders and learn more about when current orders are arriving. That gives the chatbot a role beyond product discovery, linking shopping research with order management inside the same app.
Still, the assistant has limits. TechCrunch found in tests that Rufus was a decent shopping companion and mostly avoided problematic answers to questions outside shopping. But it did not always get its facts right.
Another limitation is the assistant's connection to Amazon's own catalog. Even though that catalog is large, TechCrunch noted that being limited to Amazon's catalog could sometimes affect the quality of its recommendations. A shopping assistant that only works inside one retailer's inventory may be useful, but its answers are shaped by what that retailer sells and surfaces.
Amazon says it will continue to improve Rufus over time. For now, the launch gives all U.S. Amazon app users a new AI layer for product discovery, comparison, recommendations and order help. The practical test will be whether shoppers find that a chat-based assistant makes buying decisions faster, clearer or more trustworthy than the familiar mix of search, filters, reviews and product pages.