Here’s a number that should make any taxi driver sit up: 20,000. That’s how many driverless cars Waymo reportedly wants on the road by the end of 2027. Today the Waymo robotaxi fleet sits at roughly 4,000 vehicles. So we’re talking about five times bigger in about 14 months.
A robotaxi, by the way, is a self-driving car you hail through an app, with nobody in the driver’s seat. They’re not science fiction anymore. But going from thousands to tens of thousands is a very different game. Let’s dig into what’s going on.
So what’s Waymo planning?
The news comes from Semafor, which reported on 19 October that Waymo plans to double its fleet to 8,000 cars by the end of 2026. Then it wants to reach 20,000 by the end of 2027. The report cited “people familiar with the matter,” and Waymo declined to comment.
Think about that first step for a second. Doubling a fleet of self-driving cars in under three months is a big ask. As Gizmodo put it, the expansion would happen over the next 14 months, using newly secured capital.
Waymo, which Alphabet owns, already runs paid rides in 15 US cities. Those include Atlanta, Los Angeles, Miami and the San Francisco Bay Area. Just last week, it also opened service to everyone in San Antonio, San Diego and Tampa.
And it isn’t stopping at American borders. Waymo has already announced plans for London and Tokyo. Then, on 20 October, it added Madrid, with first riders targeted for 2027.
So this isn’t a lab experiment anymore. It’s a business trying to scale fast, and that changes the questions we should ask.
Where’s the money coming from?
Cars cost money, and lots of cars cost lots of money. In October, Waymo took on a $5 billion loan, its first ever debt financing. The lenders were PIMCO, Blackstone and Sixth Street, according to The Next Web.
Waymo called the loan an “important step in our evolution into a scaling commercial enterprise.” That’s corporate speak, sure. But it tells you something real. Debt is what companies take on when they expect steady revenue to pay it back.
That loan also comes on top of the $16 billion Waymo raised from investors earlier in 2026. In other words, the war chest is huge. So the 20,000-car goal isn’t only a dream on a slide deck. There’s actual cash behind it.
Honestly, this is the part we find most telling. A loan has to be paid back on schedule, so lenders tend to be pickier than investors chasing a big upside. When big credit firms sign up, it suggests they think the business is close to working.
Meet the Ojai, the car behind the Waymo robotaxi fleet
Most of the new cars will be the Ojai, Waymo’s next-generation robotaxi. It starts life as an electric car from Zeekr, a Chinese brand owned by Geely. Waymo then fits its own US-developed self-driving system at a facility in Mesa, Arizona.
The Ojai runs Waymo’s sixth-generation driving system. Waymo says it uses “fewer, less expensive components” and is “designed to handle a wider range of driving conditions.” Cheaper parts matter a lot here, because every dollar saved per car adds up fast across 20,000 cars.
Why not keep the old cars? Well, today’s fleet is mostly Jaguar I-Pace models on the fifth-generation system. But Jaguar stopped making the I-Pace in 2024, so Waymo needed a new ride.
There’s a trade twist, too. Chinese cars face a 100% US tariff, and federal rules restrict Chinese connected-car software. So Waymo imports the Zeekr cars without that software and adds its own tech in Arizona. The New York Post, citing research firm MoffettNathanson, reported that Waymo is on pace to import about 5,100 Ojai vehicles by the end of 2026.
Why this one’s a big deal
First, it widens the gap with rivals. In Texas, Waymo runs hundreds of cars, while Tesla had just 42 robotaxis, according to state records cited by The Next Web. A 20,000-car fleet would put Waymo far ahead.
Second, scale changes the economics. A few hundred cars in one city is a pilot. Thousands of cars across many cities start to look like a real transport network, and that’s when prices and wait times can shift.
Third, it shows how physical AI is leaving the lab. If you want the basics, our guide to physical AI explains how software learns to act in the real world. A robotaxi is probably the clearest example most people will ever ride in.
Finally, the race is going global. We covered how Uber and Pony.ai are chasing London. And in Madrid, WeRide and Uber are already ahead of Waymo, holding Spain’s first permit for driverless cars. So Waymo isn’t the only one moving fast.
Okay, but what could go wrong?
Quite a lot, honestly. In May 2026, Waymo paused service in five cities “after its cars kept driving into flooded roads.” That’s the kind of edge case that gets harder to manage as a fleet grows.
As The Next Web put it, “a bigger fleet brings bigger risks.” More cars mean more strange situations on the road and more chances for a viral mistake.
There’s also politics. Gizmodo points to “growing scrutiny from U.S. lawmakers,” and the Chinese-built base car could become a sore spot. Tariffs and software rules can change, and a plan built on imports is exposed to that.
And remember, these targets come from unnamed sources, not from Waymo itself. Plans like this can slip. Still, the loan and the factory work suggest the company is serious about the direction, even if the exact numbers move.
We’d also love more public data on safety and reliability as the fleet grows. That’s the real test, since trust is what gets people into the back seat.
What a bigger Waymo robotaxi fleet means for you
If you live in a city Waymo serves, you’ll probably see a lot more empty driver’s seats soon. If you don’t, this still matters, because whatever works at this scale tends to spread.
For emerging markets like Armenia, robotaxis aren’t around the corner. But the pieces behind them are worth watching: sensors, mapping, fleet software and remote support. Those are jobs and skills, and they don’t all need to sit in California. If you’re a student or an engineer, that’s a field worth keeping an eye on.
So the Waymo robotaxi fleet is turning into a real-world stress test for physical AI. If it works at 20,000 cars, the conversation shifts from “can robots drive?” to “who gets to run them, and where?” And honestly, that second question might turn out to be the harder one.
Here’s our question for you. Would you hop into a car with no driver tomorrow, or would you want to watch someone else go first?