Wabi, the AI agent that builds mini apps, now valued at $1.25B after latest round
- ID
- 40770
- Status
- summarized
- Published
- 29 Oct 2026, 10:35 PM
- Fetched
- 30 Oct 2026, 12:28 AM
- Provider
- TechCrunch
- Category
- technology
- Original URL
- https://techcrunch.com/2026/10/29/wabi-the-ai-agent-that-builds-mini-apps-now-valued-at-1-25b-after-latest-round/
- Source URL
- https://techcrunch.com/feed/
Summary
- Score
- 3.5
- Created
- 30 Oct 2026, 12:28 AM
- Tags
- Audience
- vibe_codersfoundersai_agent_users
What happened
Wabi, the AI agent startup founded by Eugenia Kuyda (previously of Replika), raised $55M at a $1.25B valuation, led by IVP with participation from a16z, Forerunner, Amazon Alexa Fund, and angels including Tobi Lütke, Marc Benioff, Marissa Mayer, Dick Costolo, and Alex Pall. Wabi originally let people build apps via prompts, then pivoted last month to a 'personal agent' / AI messenger model ('Wabi 2.0') that generates mini-app interfaces on demand for things like meal planning, restaurant tracking, fitness plans, and trivia games. The article gives no product pricing, API details, model choices, benchmarks, or user numbers — it is a funding announcement with example use cases.
Why it matters
The only hard facts here are the $55M raise, the $1.25B valuation, and the pivot from prompt-to-app builder to personal agent that generates UI — no pricing, API, or performance data a builder could evaluate or adopt. Treat it as a competitive signal, not a tool: if you are building prompt-to-app or agent-generated-interface products, investors are currently pricing this category at large multiples (the article cites Instinct at a $10B valuation in September and Tolan's $30M raise), so expect more well-funded entrants competing for the same 'describe it and get an app' use case. There is nothing in this text about Malaysia or Southeast Asia, so no local regulatory, payments, cloud, or funding implication can be drawn from it.
Discussion angle
Wabi's pitch is that the agent builds the interface it needs on the fly. What would you actually need to see — latency, cost per generated app, whether the generated UI survives edits, data retention — before you'd use a generated mini-app over a static form or a no-code tool? And does a $1.25B valuation on a one-month-old pivot tell us anything actionable, or just that agent funding is frothy?