Wabi Raises $55 Million at $1.25 Billion Valuation as Its App Builder Becomes a Personal Agent
Wabi has raised $55 million in a round announced on October 29, 2026, at a valuation of $1.25 billion. IVP led the round. Andreessen Horowitz, Forerunner and Amazon Alexa Fund also participated. The company said the round followed strong early demand.
The company was founded by Eugenia Kuyda, who previously founded the AI companion startup Replika. Angel investors include Shopify CEO Tobi Lütke, Salesforce CEO Marc Benioff, former Yahoo head Marissa Mayer, former Twitter CEO Dick Costolo and Alex Pall.
Wabi began as a tool that let anyone build applications from natural-language prompts, meaning plain-English instructions. That product has since been reworked. In late September, TechCrunch reported that Wabi had repositioned its prompt-based app builder as a personal AI agent capable of creating interfaces on demand. The company had also pivoted to an AI messaging experience it calls Wabi 2.0. It now describes itself as a new kind of personal AI assistant built for life together, and its homepage reads 'The AI assistant you use together.'
The published product details point to a few specific design choices. Users can add Wabi widgets to iOS or Android home screens. They can also start a session on a computer and pick it up on a phone, or the reverse. That means session state, the record of what has happened so far, has to persist across devices instead of living inside a single app session. A shared assistant used by several people adds another requirement: the agent must keep each user's context, permissions and history separate within one conversational thread.
TechCrunch calls the company an AI agent, and its own materials use the term 'personal agent.' The distinction matters in practice. A prompt-driven app builder produces an artifact, a mini-app that is generated, run and then used. A persistent agent is judged on what it remembers, how reliably it acts across sessions and whether it can be trusted with tasks that run over time. The pivot moves Wabi from the first category toward the second.
The broader context here is a crowded field. Several companies are building assistants that sit inside messaging apps, and some are trying to make generated interfaces part of the conversation rather than a separate tool. Among the angels are the chief executives of a major e-commerce platform and a major enterprise software company. The funding announcement does not describe how Wabi plans to use the capital.
In my view, the open question is less whether generated mini-apps can be produced on demand, which the original product already showed, and more whether a conversational surface that follows a user from desktop to phone to home screen becomes something people return to every day. That is an engineering and product question, and usage will answer it, not a funding round. If it works, software stops being something a person opens and becomes something a person asks for, wherever they happen to be.
