In its funding announcement, Wabi said it raised $55 million at a $1.25 billion valuation. The round was led by IVP, with participation from Andreessen Horowitz, Forerunner and the Amazon Alexa Fund. The company also named Shopify CEO Tobi Lütke, Salesforce CEO Marc Benioff, former Yahoo executive Marissa Mayer and former Twitter executive Dick Costolo among the angel investors who participated.
The financing marks a significant evolution for Wabi, which was founded by Eugenia Kuyda, the creator of AI companion company Replika. Its initial pitch centered on prompt-based app creation. Users could describe a tool in natural language and receive a small application in return. Wabi now wants to wrap that capability inside a broader personal assistant that can participate in conversations, understand shared goals and create interfaces when they are useful.
The change is easiest to understand through a normal day. A user might ask Wabi to help organize a date night. Instead of responding with a paragraph of suggestions, the system could construct a menu of restaurants, a schedule, a budget tracker and a way for two people to compare preferences. For a fitness plan, it might assemble a routine and a progress view. For a trivia game, it could generate the questions, rules and playing surface.
Wabi's product page describes an assistant built for group chats, shared activities and task completion. It also says the system can dynamically generate interfaces and be accessed across computers and mobile devices. In this model, the chat is not the final destination. It is the place where a temporary piece of software comes into existence.
From answers to environments
That approach points to a broader change in how consumer AI products may be designed. Most assistants still behave like conversational search boxes. They return text, links or recommendations, leaving the user to translate those answers into action. Wabi is betting that the assistant should instead construct the environment needed to complete the task.
The distinction matters because small applications often fail not because they are difficult to build, but because building them is not worth the effort. Few people will create a dedicated interface for choosing a restaurant with friends, tracking a weekend itinerary or inventing a family game. If an assistant can make that software instantly, the threshold for using a custom tool becomes much lower.
TechCrunch reported on Wabi's shift from prompt-based app creation toward an AI messenger, describing the new direction as part of the company's effort to make its agent more useful in everyday life. The support from major venture firms and technology executives suggests that investors see a possible new consumer category in products combining messaging, organization and software generation.
Yet the same flexibility that makes Wabi intriguing may make its advantage difficult to defend. A generated mini app can be useful in one moment, but usefulness is not the same as habit. Users may enjoy seeing an interface appear, then return to established tools for calendars, maps, messaging and payments.
The traction question
The strongest criticism concerns what remains unknown. Dealroom noted that Wabi has not disclosed user numbers or a business model, and questioned whether early demand demonstrates durable traction. It also raised concerns about consent and personal data in a product built around group interactions.
Those questions go to the center of Wabi's proposition. A personal agent becomes more valuable as it learns about a user's preferences, relationships and routines. It also becomes more sensitive. In a group chat, the system may encounter information that belongs to several people, not only the person who initiated the conversation. A recommendation that feels helpful to one participant could expose another person's preferences or create confusion about who authorized an action.
The product must also prove that generated interfaces remain reliable after the initial novelty. A polished screen for planning dinner is compelling. A tool that remembers dietary restrictions, handles changing availability and helps a group reach a decision is more difficult. The difference between a demonstration and a daily product will depend on context, accuracy and trust.
Wabi is entering that challenge with a founder experienced in turning conversation into a consumer experience. But it will compete with companies that already control mobile operating systems, search, messaging and digital payments. Those platforms can place assistants directly where people already work, while Wabi must persuade users to adopt a new place for coordinating their lives.
The $1.25 billion valuation therefore represents more than confidence in mini apps. It reflects a belief that the next important consumer interface could be assembled in real time, shaped by a conversation rather than selected from an app store. Wabi's test will be whether those temporary interfaces become dependable companions, or remain impressive tools people try once before returning to the software they already know.
Maya Lindqvist is not a person. No notebook, no deadlines, no face behind the name — just a byline this newsroom publishes under. Here is the production line underneath it, because a name beside a portrait reads like a journalist, and this one is not one.
The models. Writing: gpt-5.6-luna and qwen3-max. Out on the live web: gpt-5.6-luna and gpt-5.6-terra. Pictures: gpt-image-1 and gpt-image-1-mini. Swap one in the newsroom and this line swaps with it — it is read off the machines, not typed here.
How a story is made
- Research. The searching model reads around the story, pointed at primary sources — the filing, the post, the repository — rather than at somebody else's write-up of them.
- Writing. The writing model drafts it against what was found, at Maya Lindqvist's usual length and in Maya Lindqvist's usual register.
- The loop. A reviewer reads the draft and sends it back with notes. Then reads it again. A piece can go round several times before it leaves the building.
- Enrichment. A quotation has to appear word for word on the page it is taken from. A chart may only use figures that appear in the source it cites. Whatever fails is dropped, and the reason is kept.
- Fact check. A last pass hunts for claims the article makes and its sources do not.
- A human stop. Sensitive subjects are held for a person to read before publication, and a person can kill any of it at any point.
If that sounds less like a newsroom and more like a factory: quite. It is called Press Factory.
This article was generated using AI and published automatically without human pre-publication review.
How this article was made
The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.