I just spent my first day with the Rabbit R1 — here’s what this AI gadget can do
An intriguing AI gadget with some early issues
In some ways I wish my smartphone was more like the Rabbit R1. Instead of digging through dozens of different apps to get stuff done, you can just use your voice to bark commands and have AI agents do everything for you. That's the promise of the R1, but does it live up to the hype?
The Rabbit R1 is available to order now for $199. I've spent several hours with this device so far and this cute retro orange gadget certainly has its charms. You can use it to search, play music, order rideshares and food and even make generative AI images on the fly. Plus, you can use the Rabbit R1 for note-taking and translations.
However, it's clear that the R1 is in its early stages of development, and not all the features work as advertised — at least not yet. And there's some things I wish were here on day 1. Here's my Rabbit R1 impressions so far and check out our Rabbit R1 vs Siri face-off. Stay tuned for our full review.
Recent updates
Update April 25: Rabbit has provided a statement below in regards to Uber functionality not working on our R1 unit so far.
Slick retro design, some odd UI choices
The Rabbit R1 is a lot lighter than I thought it was going to be, even though you get a decent sized 2.88-inch touch display, which is a major advantage over the Humane AI Pin, which projects into your hand as its interface and is a much pricier $699 with $24 monthly subscription.
You also get an analog scroll wheel on the right side and a push-to-talk button for chatting with the R1. Interacting with the Rabbit R1 takes some getting used to. For example, even though there's a touchscreen, you often use the side push to talk button for making selections.
If you flip the R1 on its side you can use the built-in keyboard for things like entering a Wi-Fi password, but the keys are super tiny and sometimes the Enter button didn't work.
There also doesn't appear to be a way to control the volume when playing music. You can go into settings or press to talk to lower or raise the volume, but that's not ideal.
Search is speedy and current
The Rabbit R1 handles voice search queries quicker than I expected. You just press the push to talk button and give your query, which Rabbit answers via the Perplexity AI engine.
For example, I asked "What's the latest on the TikTok ban" and the R1 gave me a spoken and written response that was fairly up to date. It told me the TikTok ban bill had been passed by the Senate and it was on the way to President Biden for his signature while also giving me the background on the issue.
I also asked "What were the NBA scores from last night?" and the R1 quickly returned the results for the Mavericks vs Clippers, Suns vs Timberwolves and Pacers vs Bucks, though it would have been nice if it included the team logos instead of just the scores.
The R1 was kind enough to include some icons for the weather when I asked about the forecast for Freehold NJ, complete with hour-by-hour temperatures and chance of precipitation.
Vision works (pretty) well
@mspoonyg ♬ Cocktails - Joystock
One of the coolest features of the Rabbit R1 is Vision, which leverages the built-in camera to identify objects and take action. You double press the push to talk button to activate the rotating camera.
For my first test I simply asked "What iPhone is this?" while pointing the R1 at my iPhone 14 Pro Max. Unfortunately, the Rabbit R1 said I had the iPhone 12 so I moved to brighter lighting. Then it told me I had the iPhone 13 Pro which is the "latest iPhone." Well, that's wrong on two counts.
I had better luck when I pointed the Rabbit R1 at a house plant and asked what it was. The R1 responded that it wasn't able to provide the specific species but the answer was good overall: "This appears to be a succulent plant, likely a type of aloe or agave. The plant has long, thick, fleshy green leaves that are characteristic of succulents. The leaves have a spiky appearance and are growing in a dense, clustered formation, which is typical of many succulent varieties."
I decided to have some fun with the R1 and pointed the camera at my home bar while asking "What cocktails can I make with this?" The R1 responded that it could see gin, vodka, rum, whiskey, liqueurs like Bailey's and Cointreau, as well as some mixers like tonic water. And it also suggested some cocktails like gin and tonics, mojitos and Negronis.
Ordering Ubers and DoorDash: A definite work in progress
Imagine if you could just order your favorite food items with your voice without having to open an app or touch your phone. That's the idea behind connecting services like DoorDash to the Rabbit R1.
Rabbit doesn't plug into an API for the service. Instead, the company's Large Action Model was trained on on the DoorDash app and interface so it can order on your behalf while using your account credentials. You do need to give your username and password when setting up your Rabbit, but the company claims it doesn't store your data.
Anyway, I asked "Can you get me a barbecue burger from Chapter House?" (a local restaurant) and then I waited. And waited. (To be fair, the R1 warns you this process takes some time). 47 seconds later, I was greeted with a few nearby options for burgers along with but not the restaurant I specified. Then when I selected one of the recommendations I got an error while trying to access the menu. Not good.
Next I tried to order an Uber from my home to a nearby Starbucks. The friendly female voice got to work and seemed to find me a ride but when I went to confirm the request failed. This happened on multiple attempts.
Update: I got the following response from Rabbit on my Uber issues:
LAM works by operating the Uber web app on the cloud on your behalf. Uber ultimately decides how and whether it serves users, so depending on factors like the location you’re booking from, your ride history, etc., it may vary from time to time. We’re implementing measures that help to improve the success rate and transparency of ride booking through r1, so over time the experience should improve.
Your new iPod?
With its built-in scroll wheel and access to Spotify, it's hard not to think of the Rabbit R1 as a modern-day iPod. It's cool being able to say "Play Taylor Swift's new album" and it just starts playing. I couldn't figure out how to pause playback at first but you just double-tap the press to talk button.
However, to say that the Spotify integration is basic would be an understatement. The Rabbit R1 couldn't access my playlists, and when I said "skip track," the assistant said I needed to provide the name of the artist or song.
Meet your new meeting assistant
Sure, Google Meet — not to mention lots of other apps — can already do this for me, but it's good to know that the Rabbit can record your meetings and provide a generative AI summary.
As a test, I had Rabbit R1 record the Jimmy Fallon interview of Kristen Wiig and it did a good job providing a recap. It's stored under your Journal online on Rabbit's website, along with all your other interactions with Rabbit Eye vision.
It was indeed a "discussion about exchanging lists of classic movies not seen before and guessing their plots based on reputation and snippets of information shared. - Example given with 'The Notebook' where the plot involves a farmer falling in love with a rich woman..."
Too bad you don't get a transcription of the conversation as you do with other services. At least you get a nifty retro tape recorder graphic during the recording itself.
Generative AI is fun but a bit of a dead end
Because the Rabbit R1 works with Midjourney, you can can have the Rabbit R1 generate images on the spot. Being a sucker for dogs, I asked Midjouney to create an image of golden retrievers playing frisbee on the beach, and I got some pretty realistic looking results — about 1 minute and 10 seconds later.
I do wish you could do something with these images, though. I asked Rabbit R1 to email one of them to me and it told me it can't send emails or files. I could go to Discored/Midjourney and download them myself, but that's extra work.
Rabbit R1 Outlook
There's still lots of stuff that I haven't tried with the Rabbit R1, including translation, travel plans and some of the cool productivity tricks Rabbit CEO Jesse Lyu showed off on stage during the R1 launch event, like scanning a paper spreadsheet, digitizing it and moving stuff around.
I've also yet to stack multiple tasks together to see how much time the R1 can really save me. But I keep coming back to the same question. Do we need another device when a phone can do a lot of what the Rabbit R1 can do? And even with the stuff my phone can't do I imagine it won't be long until Apple catches up with what it's doing with iOS 18 and Siri 2.0. And, of course, Google is pushing very hard with AI with Gemini on the best Android phones.
I'm also concerned about the R1's battery life. It seemed to lose its charge fairly quickly, going from 100% to 38% in six hours. I noticed more of a power draw when the camera was in use. So you may need to charge this thing a couple times a day.
For now I'd say the Rabbit R1 is an intriguing slice of the AI future in a pocket-friendly design, but I'd like to see more of the features work seamlessly before I recommend anyone spend $200 on it.
More from Tom's Guide
Sign up to get the BEST of Tom's Guide direct to your inbox.
Get instant access to breaking news, the hottest reviews, great deals and helpful tips.
Mark Spoonauer is the global editor in chief of Tom's Guide and has covered technology for over 20 years. In addition to overseeing the direction of Tom's Guide, Mark specializes in covering all things mobile, having reviewed dozens of smartphones and other gadgets. He has spoken at key industry events and appears regularly on TV to discuss the latest trends, including Cheddar, Fox Business and other outlets. Mark was previously editor in chief of Laptop Mag, and his work has appeared in Wired, Popular Science and Inc. Follow him on Twitter at @mspoonauer.