- Rabbit r1 announces web-based AI agent update
- CEO demonstrates expanded capabilities across various online tasks
- Update aims to revive initial product promises
Second time’s the charm?
Rabbit’s r1 device is set to receive a major update on October 1, introducing a web-based ‘large action model’ (LAM) agent.

This development aims to fulfill the company’s initial vision of a versatile AI assistant capable of interacting with various web services.
From hype to reality
CEO Jesse Lyu acknowledges the company’s overly ambitious start but remains confident in the r1’s potential.

The new update promises to expand the device’s capabilities beyond its current limited set of services, enabling it to navigate any website perform a wide range of tasks.
Putting it to the test
During a demonstration, the updated LAM showcased its ability to break down complex tasks into steps execute them by analyzing on-screen elements.
From concert ticket purchases to domain registrations, the AI agent displayed promising versatility adaptability across different web interfaces.