Remember when the Rabbit R1 launched in April 2024 and everyone lost their minds? The little orange AI gadget that was supposed to replace your phone? I was skeptical. I mean, we’ve been promised “the next smartphone killer” for years — Google Glass, Amazon Echo Frames, Humane AI Pin — and they all flopped. But the Rabbit R1 was different. It had a physical scroll wheel, a tiny screen, and this weird retro charm that made you want to believe. I ordered one on day one. Three months later, here’s the honest truth.
The Rabbit R1 is not going to replace your phone. Not even close. But it’s found a surprising place in my life, and I think it points to where AI hardware might actually go.
What It Does Well: The “Do Anything” Promise
The core idea of the R1 is that you train it to perform tasks across apps — booking an Uber, ordering food, sending a message — all by speaking to it naturally. The company calls it a “Large Action Model” rather than a language model. In theory, you say “Book me a ride to the airport at 7 AM tomorrow” and it just does it. No app switching. No tapping. And honestly? When it works, it’s magical. I’ve used it to order pizza from Domino’s, set reminders, and even play music on Spotify. It’s faster than pulling out my phone if I’m driving or cooking. The scroll wheel is satisfying to use — it clicks with a tactile feedback that reminds me of the old iPod click wheel. That alone makes it more fun than any smart speaker.
Where It Falls Short: The Rough Edges
But here’s the catch: it only works about 70% of the time. I’ve tried to order from Uber Eats and it misunderstood the restaurant name three times. I asked it to “send a text to Mom” and it sent it to my friend Mark instead. Awkward. The screen is tiny and low-res — 2.88 inches, 240×536 pixels — so reading anything is a chore. Battery life is about 5 hours of active use, which is mediocre. And the speaker is tinny; music sounds like it’s coming from a can. For $199, it’s fun, but it’s not polished.