Happy First Sunday of Summer! Or, to a certain subset of humans, Happy Tesla Robotaxi Launch Day. However, everyone — people, pets, anything that breathes — should be paying attention to these autonomous cabs, especially if you live in Austin, Texas, where the self-driving rideshare service is being deployed.
This is a Tesla first. Most importantly, this is a Tesla finally. After (counts fingers) six years of saying “next year,” CEO Elon Musk can now say, “Today.” Musk first announced his vision for the Robotaxi (the service’s official name) during the company’s Autonomy Investor Day in April 2019. And in what has become a bad habit, a nervous twitch, or an uncontrollable reflex, Musk said robotaxis would be on the road in 2020. As CNBC reported at the time:


“‘I feel very confident predicting autonomous robotaxis for Tesla next year,’ Musk said on stage at the Tesla Autonomy Investor Day in Palo Alto, California. They won’t be ‘in all jurisdictions, because we won’t have regulatory approval everywhere, but I am confident we will have at least regulatory approval somewhere, literally next year,’ he said.”
Um, just kidding? The year 2020 came and, mercifully, went (it was a peak viral time, ‘member?). But then, so did 2021, 2022, and 2023 come and go without a self-driving Tesla taxi in sight. Then Musk threw a curveball at our heads with the Cybercab, something he announced during the We, Robot event last October. I say curveball because, yes, although the Cybercab is a robotaxi, it’s not the one we’re actually getting. Not yet, anyway.
Prelims: Cybercab vs. Robotaxi
The Cybercab and robotaxi are not equal parts of the same name. There is a difference between the two, which are often erroneously interchanged. The Cybercab is just another autonomous vehicle (AV) concept (important keyword right there) that has no steering wheel or pedals. A robotaxi, however, can be any make or model of vehicle that has been modified to operate without a driver.
For example, Waymo launched its self-driving service with a Chrysler Pacifica. Eventually, it switched to the Jaguar I-Pace before inking partnership deals with Hyundai, Zeekr, and Toyota. Other mass-produced vehicles converted into robocabs include the Chevy Bolt, by the now-defunct Cruise, as well as the Ford Fusion and Volvo XC90, which Uber used for its self-driving fleet before selling off the whole operation. Amazon-owned Zoox is building its own bespoke van-like self-driving shuttles in an all-new manufacturing facility with public service to start soon in Las Vegas and San Francisco.
Main Card: Tesla vs. Everyone
Or, in this case, really just Waymo for now. Multiple startups and big brother-backed robotaxi services have crashed (literally) and burned out of existence. Or sued out of operations.
Cruise and Uber autonomous vehicles were involved in pedestrian collisions, some resulting in fatalities. General Motors, a 90% majority owner of Cruise at the time, shuttered the autonomous ride-hailing service before wholly buying out the company earlier this year in order to integrate the tech within its vehicle portfolio.
As for Uber, which graduated beyond peer-to-peer ride-hailing, the multi-level transportation company is not completely out of the robotaxi market. Uber has announced partnerships with autonomous vehicle tech companies, May Mobility and Momenta, to scale up its robotaxi offerings in the U.S. and Europe, respectively. Although expected to launch in Austin by the end of this year, autonomous rides won’t be broadly available until 2026. Same for Lyft, which also buddied up with May Mobility and other AV/ADAS companies Mobileye and Nexar. Its autonomous cab fleet is set to deploy sometime this year, starting in Atlanta.
Until then, there’s Waymo, which, the record will show, was the first company to launch a commercial ride-hailing robotaxi service. Those paid rides kicked off in Phoenix in 2018, a year before Musk even mentioned the idea of such a thing. Formerly under the Google umbrella, Waymo has since gone independent and currently operates in multiple cities, including Los Angeles, San Francisco, and, you betcha, Austin. Atlanta and Miami are next on the list.
Main Event: Musk vs. Musk
I’m not even going to sugarcoat this: Musk needs a win here. Of all the promises and intent and speculation and setbacks, the robotaxi scheme, er, service has to work. Because if you haven’t been following the spectacle, Tesla sales are tanking globally, the haloed Cybertruck has become a Cyber-not, and Musk, a once favored friend with benefits to the U.S. presidency is now a frenemy at best.
But Musk is as excited as can be about today’s Robotaxi launch, which, by Tesla standards, has been rather low-key. From the announcement (psst, here’s some news) to the invites (“select guests”) to the fleet size (about a dozen Model Ys), the kickoff is more of a soft opening. As the Associated Press reports:
“The test is beginning modestly enough. Tesla is remotely monitoring the vehicles and putting a person in the passenger seat in case of trouble. The number of Teslas deployed will also be small — just 10 or 12 vehicles — and will only pick up passengers in a limited, geofenced area.”
Hmm, so, although there is no actual driver, there is a co-driver, so to speak. Now, whether this is Tesla’s own protocol or a requirement of its self-driving fleet permits with the city is unknown without me digging through Austin’s AV regulations.
The geofencing does come off as a little suspicious, though. Is controlling the service area to minimize issues on launch day and before full expansion? Tesla hasn’t stated what neighborhoods are good-to-go, but it appears that once you’re in the dedicated Robotaxi app, you’ll see a map with distinct borders, as shown in this Teslarati article. You can also see the app in action courtesy of Tesla podcaster Rob Maurer:
Here’s a quick overview of the Tesla Robotaxi app! pic.twitter.com/AAtPZS1E5T
— Rob Maurer (@TeslaPodcast) June 22, 2025
Note that access to the Tesla Robotaxi app is an invite-only deal at the moment, so you can’t request a self-driving Model Y cab without the coveted “Early Rider access.” Should you receive the golden ticket, Tesla is very clear about what its Robotaxi service currently offers. Sawyer Merritt, a Tesla investor and tech news source with 900,000 Twitter/X followers, shared his invite on the social media platform:
BREAKING: I have been invited to ride in Tesla’s Model Y Robotaxis on day one this Sunday in Austin, Texas!
IT’S GO TIME BABY!!! pic.twitter.com/FGzG8oW9El
— Sawyer Merritt (@SawyerMerritt) June 20, 2025
Merritt was also allowed to share the parameters of the Robotaxi service, the privacy rules (e.g., the vehicle cabin camera will not be recording), and what costs a Tesla robocab might include. Just for today, though, Robotaxi rides cost less than five bucks. Hey, thanks, Elon:
The @Tesla_AI robotaxi launch begins in Austin this afternoon with customers paying a $4.20 flat fee!
— Elon Musk (@elonmusk) June 22, 2025
Decision: A Draw
Look, this bout should’ve happened years ago. Tesla, past its prime in terms of favorability, is late to the robotaxi market. The bright side is that those who experienced a Tesla Robotaxi ride in Austin offered nothing but praise, to which skeptics will respond, “Of course, when you select who gets a seat.” And it’s a backseat at that. Nevertheless, for what it’s worth, the Early Riders have been livestreaming their entire commute, warts and all. You can watch the first Robotaxi guests via Tesla’s feed:
First @robotaxi experiences in thread below
— Tesla (@Tesla) June 22, 2025
Many of the Early Riders have also ridden multiple times. And, knock on wood, so far so good! No navigation glitches have been reported. Even better, no collisions. For that, Musk and Tesla get a win. But it’s also a very limited group traveling in a very limited space.
When Waymo launched its Waymo One rideshare service to the public, its ridership was also limited to approved applicants. However, riders could travel throughout a greater area, including cities like Chandler, which is 25 miles away from downtown Phoenix. The Tesla Robotaxi maps that have been shared don’t look as expansive. Oh, and 100 minivans were in operation and not, you know, 10.
To get a knockout, Tesla will need to ramp up its Robotaxi fleet, even if just within Austin. But with Musk shifting his focus back on his EV company, maybe that might happen. In an interview with CNBC, Musk said:
“My prediction is that probably by the end of next year, we’ll have probably hundreds of thousands, if not over a million Teslas doing self-driving in the U.S.”
Flabbergasting. Even the CNBC host was at a loss for words. How Musk sees this happening is by using Tesla owners. The plan, which you may have heard of, will be “a model which is some combination of Uber and Airbnb” where owners can “add or subtract” their car from the rideshare fleet.
All of the above is ambitious, but what is Tesla (and Musk) without fantastic, far-reaching goals? To the doubters, what Musk said in 2019 still holds true today:
“Sometimes I am not on time, but I get it done.”
Top graphic images: Tesla; Carolco/TriStar
Still no LIDAR or apparently even radar? Yeah, I’ll be steering clear. Apparently it’s cheaper, but I wonder if that’s still true if you include the lawsuit when (not if) it makes an error and kill someone.
I have an idea about the geofencing.
Teslas rely on cameras to evaluate the objects around them. Several tests have identified weaknesses in their camera-only system
I bet they 3D scanned the roadways in the service area like iRacing does with race tracks. They used the data to create a model of the service area that is used to navigate around the limited area.
If that is a thing they may have done, it’s not going to scale easily. I assume it takes several passes on different days for the scan to get a clear picture of all the fixtures near the roadway, since parked vehicles would be blocking curbs and other roadside obstacles.
Considering Tesla can’t even automate their Las Vegas disco tunnel, I’m highly doubtful of their ability to do with their cheap camera based system what Waymo struggles to do with a far more advanced system.
I wonder how long it’ll be before we see these flipped on their sides or dumped in canals?
Austin was a terrible choice to launch this considering the demographics in the city. There are so few places that would be as hostile to him as Austin that he would have to assemble a team to try to find them. Did he think that because it’s the capitol of a red state that it would just automatically be accepted? Or was this decided back when he was still the benevolent Techno Paragon and people were still using the slogan “Keep Austin Weird” for tourism?
Why not launch this in Phoenix Arizona where they’d not only be more friendly, but the streets are less snarled and it would be easier to maintain the cars because Phoenix never gets any rain?
He moved Tesla to Austin a few years ago and employs 20k people there. I assume he has clout in the area – at least enough to push through approval to set these things loose.
The Mayor of almost any other city would get an easy political win by making it difficult for Elon.
Will the inevitable traffic incidents (hopefully, no fatalities) be investigated by RoboCop?
Waymo, the current leaders in true driverless taxi operations, is making billions on them.
Wait, no..they’re not…
How this ever replaces a business model (Uber/Lyft) where a person literally offers their own car, labor and covers all the running costs is beyond me.
Someone still has to own these things, clean up the barf in the seat, repair them etc.
As with anything Musk related, I’ll pass even if these were in my area.
I do think its funny that most of these robotaxi operations are in fairly clear climate areas, especially Tesla, that somehow with their camera only operation they’re going to take over the world. Waiting for them to try it in Chicago in January or Seattle in…any random Tuesday I guess.
At least Waymo is trying for NYC, they have to get permitting but that’s a bold move, NYC winters are no joke.
Just a reality check from Nov. 4th 2022;
https://www.autonews.com/mobility-report/autonomous-vehicle-reality-check-after-160-billion-spent/
Now how many abandoned regional rail lines do you think could be rehabilitated to modern standards for that, and its benefit to traffic and pollution reduction?
For relatively flat land and basic track, expect costs between $250,000 and $1 million per mile, according to Compass International and FreightWaves.
$4.20 for a trip sounds great, but I’m waiting for the Special $1K for 100K miles discount pass.
” To the doubters, what Musk said in 2019 still holds true today:
“Sometimes I am not on time, but I get it done.””
I must have missed all the “New York-DC in 30 minutes” Hyperloops, car elevators, Roadster 2.0 with SpaceX rocket thrusters, 5 minute battery swap stations, rail-beating semis, coast-to-coast car summoning, solar powered superchargers, amphibious Cybertrucks, brake pads that never need to be replaced, and SpaceX rockets between New York and London.
It’s like the inverse about the old saying for economists: They’ve correctly predicted 19 of the last three recessions.
All are just fancy ways of saying a broken clock is right twice a day.
So who ends up in court when it parks under a firetruck at 80mph.
I think I’d rather have Jason Torchinsky as a cab driver.
They are my Taxi of Tomorrow
When reached for comment Robotaxi responded with, “Please place the items in the bagging area.”
I don’t care how many frozen yogurt samples they pass out, I’m not buying. It took five years for this virus to show symptoms.