Wabi Shifts to Agent‑Centric Messenger That Generates Apps On Demand

Wabi, formerly known for letting users spin up full applications from natural‑language prompts, has announced a strategic pivot. The new release, dubbed Wabi 2.0, reimagines the product as a personal agent that converses with users and instantly builds the UI components needed to complete a task.

Instead of a traditional chat‑only experience, the messenger surfaces interactive widgets — such as a calorie tracker, a workout log, or a family calendar — directly inside the conversation thread. Each widget is generated on demand, persisted alongside the relevant chat, and can be shared with collaborators.

This approach places Wabi in direct competition with agent frameworks from Meta, OpenAI, and others that blend dialogue with productivity tools. By keeping the interface context‑aware, the startup aims to avoid the “endless scroll” problem that plagues pure chat agents.

Why it matters for GPU / AI infrastructure

On‑demand UI generation requires rapid model inference and lightweight sandbox execution, both of which increase demand for low‑latency GPU compute. As more consumer‑grade agents adopt this pattern, cloud providers like AiGpu will see higher utilization of inference‑optimized instances and a growing need for elastic scaling.

Wabi 2.0 is currently invite‑only, with codes distributed via the company’s social channels. Early testers report a smoother workflow for personal productivity and small‑team coordination.