Wabi, the AI startup known for allowing users to build apps via prompts, is shifting its product strategy to launch a messaging-based experience. The pivot comes as demand for AI agents grows, with the company aiming to position itself as a hybrid of conversation and on-demand software creation.
What Happened
Wabi announced this week that it is releasing Wabi 2.0, which founder Eugenia Kuyda describes as a “personal agent that does stuff for you and builds the interface you need in the moment.” The new product functions as an AI messenger capable of generating apps to assist with specific tasks. Kuyda, who previously founded the AI companion startup Replika, stated that the company is building “the OS for the agentic era,” where software adapts to the user’s life.
The update marks a departure from Wabi’s previous positioning against vibe-coding tools. Instead, Wabi 2.0 now competes more directly with AI agents that integrate conversations and productivity tasks within a single interface. In a post on X, Kuyda argued that chat-only agents suffer from poor usability over time, noting that “your history and ongoing tasks end up buried in one endless scroll.” She emphasized that users prefer to “tap-tap-tap, scroll, and look” rather than just type, suggesting the most powerful agents will combine conversational AI with traditional software interfaces.
Examples of potential use cases shared by the company include creating a calorie tracker, a weightlifting log, or a family calendar for kids’ activities. The app is currently available only by invite code, which is being distributed on X.
Why It Matters
This pivot reflects a broader industry trend where AI developers are rethinking how applications interact with users. While early generative AI tools focused on code generation or simple chat, the current market is moving toward agents that can execute tasks and generate custom UIs on the fly. This approach aligns with similar moves by major players; for instance, OpenAI recently announced updates to ChatGPT’s plug-ins to allow interactive experiences within the chatbot’s interface, pinning app names for better visibility.
By combining messaging with dynamic app creation, Wabi is attempting to solve the limitation of static software in a personalized context. The strategy suggests that future consumer AI may not rely on pre-built applications but rather on agents that assemble tools as needed. This shift challenges the traditional model of app stores and standalone software, positioning conversational agents as the primary interface for digital life.
The Bottom Line
Wabi has launched Wabi 2.0, an invite-only AI messenger that generates custom app interfaces on demand. Founder Eugenia Kuyda positions the product as a solution to the limitations of chat-only agents, aiming to compete in the growing market of integrated AI productivity tools.