How far ahead is Waymo today?

Is Waymo really the biggest robotaxi service today?

Yes. Waymo is the biggest robotaxi service in the U.S., and on the latest reported figures it is also ahead of Baidu on weekly ride volume worldwide.

The U.S. gap is large. Waymo said it gave more than 250,000 paid trips per week in May 2025. By March 2026, TechCrunch reported roughly 500,000 paid weekly trips. In September 2026, when Waymo opened paid rides in Denver, San Diego, and Tampa, the company again put the figure above 500,000 a week, with a stated goal of 1 million weekly rides by the end of 2026. That goal is a target, not a result.

Tesla is not in the same operational category. Texas fleet filings reported in September 2026 showed about 420 Tesla vehicles registered for its robotaxi service, but independent trackers counted only a few dozen running without anyone on board. Tesla does not publish paid ride counts.

Zoox began charging fares in Las Vegas in August 2026. It has not disclosed ride volumes or a current fleet count. That is a real commercial start, not scale.

Baidu was the global caveat, and the caveat has weakened. Apollo Go gave 3.2 million fully driverless rides in Q1 2026, with weekly rides above 350,000 in March. After the Wuhan outage on March 31, Baidu said Q2 2026 rides fell to about 1 million, which is under 80,000 a week on average. Baidu said operations in affected cities were resuming in August, so the gap may narrow again.

That puts Waymo about two to three years ahead of Tesla and Zoox in U.S. robotaxi scale, and clearly ahead of Baidu on current volume.

Is Waymo still ahead on city rollout in 2026?

Waymo is ahead on U.S. city rollout because it has turned expansion into a repeatable operating playbook.

The rhythm sped up in 2026. In February, Waymo began admitting public riders in Dallas, Houston, San Antonio, and Orlando on the same day. In July, it started employee-only driverless rides in Denver, San Diego, Las Vegas, and Tampa. On September 1, 2026, Denver, San Diego, and Tampa opened paid rides to the public. In October 2026, Waymo said it had launched its fifteenth U.S. city in September, and it opened San Antonio, San Diego, and Tampa to all riders without a waitlist.

That matters because each city is a new autonomy test. It adds different road layouts, pickup behavior, weather, construction patterns, school zones, local politics, and rider expectations. A company that can expand from one geography to many is scaling field operations, not just software.

Tesla’s rollout is younger. By July 2026 it listed six markets, Austin, Dallas, Houston, Miami, Orlando, and Tampa, with about a dozen or fewer cars active in each according to third-party trackers. Zoox has paid rides in Las Vegas and free rides in San Francisco, Austin, and Miami. Cruise, once the closest U.S. rival, fell out of the race after GM shut down its robotaxi program.

Waymo’s real advantage is the launch rhythm: map, test, engage regulators, open to employees, open to selected riders, open to all, grow the service area, add demand channels, then add fleet supply.

That makes Waymo roughly 18–30 months ahead of the closest U.S. competitors on multi-city commercial rollout. Against Baidu, the city count is less comparable because Apollo Go has deep China coverage, but China’s permit slowdown has frozen much of its new-city growth at home.

Is Waymo’s safety lead still holding up recently?

Yes. Waymo still has the strongest public safety evidence in robotaxis, but 2026 also showed the limits of that lead.

The strongest evidence is the dataset. In September 2026, Waymo updated its Safety Impact Hub with more than 270 million rider-only miles through June 2026 across Atlanta, Austin, Los Angeles, Phoenix, and San Francisco. The company reports 82% fewer injury-causing crashes and 95% fewer serious-injury-or-worse crashes than human benchmarks in the same areas. Waymo estimates this avoided 841 injury-causing crashes; that number is a company estimate built on a conservative baseline.

Waymo also benefits from unusual transparency. A 2025 academic paper on 56.7 million rider-only miles found lower crash rates than human benchmarks for serious outcomes, and its safety work has involved Swiss Re, the University of Michigan Transportation Research Institute, and Virginia Tech Transportation Institute. Few rivals publish anything comparable.

But the negative signals matter too. In January 2026, the NTSB opened an investigation after Austin ISD reported several cases of Waymo vehicles passing stopped school buses; its March preliminary report said a remote assistance worker had told one car it could proceed. In early 2026, NHTSA opened a preliminary evaluation after a Waymo vehicle struck a child near a school in Santa Monica. Both cases are still open, with no findings published. In April 2026, Waymo issued a software recall after a vehicle entered a flooded roadway in San Antonio. In May 2026, Waymo paused freeway rides over its behavior around construction zones, then recalled nearly 3,900 robotaxis in June. Freeway rides began returning in Phoenix on July 29, 2026, and reached Austin in September.

Baidu now offers a contrast. It reports about one airbag deployment per 14.4 million kilometers, but the Wuhan stall showed that a fleet-wide system failure can matter more to regulators than crash averages.

Waymo is probably 24–48 months ahead of U.S. robotaxi rivals on safety proof. The next frontier is rare, socially sensitive failures: school buses, children, floods, emergency scenes, and construction zones.

Is Waymo only winning in easy geofenced areas today?

No, but it has not solved every hard driving condition either.

Waymo’s operating domains are no longer trivial. It runs in Phoenix heat, dense San Francisco streets, Los Angeles traffic, Austin and Atlanta through Uber, Miami and Orlando, several Texas cities, Nashville, and now Denver, where winter weather will be a real test. In Denver and San Diego, riders can only book the new Ojai vehicle.

The mileage base reinforces that. More than 270 million rider-only miles through June 2026 is a different evidence base from a small pilot with a few dozen cars.

But the edge cases are not theoretical. Flooded roads caused a recall. Construction zones caused a two-month freeway pause. School-bus behavior triggered investigations. Those are the long-tail failures that separate “works most of the time” from “can be trusted as infrastructure.”

Tesla and Wayve still have a long-term argument here. Their more general learning approaches could adapt faster across geographies if they work safely. But today, neither has Waymo’s mix of rider-only mileage, paid passengers, and repeated city launches.

Waymo looks 18–36 months ahead in deployed urban complexity, but the gap is smaller, maybe 6–18 months, in specific edge cases like heavy weather, road closures, construction, and emergency scenes.

Is Tesla closing the gap on Waymo with its Texas robotaxis?

Tesla is moving faster than it was, but it is not closing the gap in a way that changes the ranking yet.

Tesla’s 2026 progress is real. Its Texas registered fleet grew from 42 vehicles in late May to about 420 by early September 2026, including dozens of Cybercabs, according to filings reported under a new Texas disclosure law. In August 2026, Nevada gave Tesla a commercial permit with a ceiling of 5,000 vehicles over 12 months; Tesla’s own Cybercab chief engineer called 5,000 a ceiling, not a target. On its Q2 2026 call, Tesla said its fleet had driven more than 380,000 unsupervised miles.

The sources do not agree on what that fleet does. Registration filings count about 420 Teslas in Texas, while third-party trackers saw as few as 20 running without a safety driver in Austin during some weeks of mid-2026. For comparison, Waymo had 577 vehicles in Texas alone in late May 2026.

The safety evidence gap is also wide. Waymo has hundreds of millions of rider-only miles and published benchmark comparisons. Tesla’s unsupervised mileage is in the hundreds of thousands, and it redacts the narrative parts of its crash reports to NHTSA. Elon Musk has tied large-scale unsupervised deployment to a future software release, FSD v15, which he placed in late 2026 or early 2027.

Tesla’s bull case stays the same: a cheap camera-only vehicle built at scale could expand faster than Waymo’s fleet. But that remains a scaling plan, not the current market.

So the gap today is still about 30–48 months in favor of Waymo.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Does Zoox’s paid Las Vegas launch put pressure on Waymo?

Some, but not much yet. Zoox is now a paid service, while Waymo still has far more volume and cities.

The federal hurdle is gone. In late July 2026, NHTSA granted Zoox a temporary exemption from eight federal safety standards for its vehicle with no steering wheel or pedals. The exemption lets Zoox charge fares and caps its commercial fleet at 2,500 vehicles a year for two years. That is a cap, not a fleet. In August 2026, Zoox started charging riders in Las Vegas, with prices it compared to a premium ride-hailing tier.

Zoox’s vehicle is still the purest robotaxi design on the road: bidirectional, with facing seats, built for autonomy from day one. It also gives free rides in San Francisco, Austin, and Miami. But California still requires state driverless deployment permits before Zoox can charge there, and Zoox has not disclosed ride counts or fleet size.

Waymo is narrowing the design gap from the other side. Its Zeekr-based Ojai is now the only vehicle riders can book in Denver and San Diego, and a refreshed silver Ojai entered testing in October 2026, with rides planned in San Francisco, Los Angeles, and Las Vegas, Zoox’s home market.

So Waymo is still about 24–36 months ahead of Zoox commercially. Zoox keeps a 12–24 month edge in purpose-built form factor, but that edge is shrinking as Ojai scales.

Has Baidu fallen behind Waymo after the Wuhan robotaxi outage?

For now, yes. Baidu remains Waymo’s strongest global rival, but the Wuhan outage cost it a quarter of growth while Waymo kept scaling.

The numbers show the break. Baidu said Apollo Go gave 3.4 million fully driverless rides in Q4 2025 and 3.2 million in Q1 2026. On March 31, 2026, around 100 Apollo Go vehicles stalled in Wuhan, stranding passengers; some reports said nearly 100, others more than 100. Chinese ministries ordered safety reviews, and reports said new Level 4 permits were paused. On its Q2 2026 call in August, Baidu said Apollo Go gave about 1 million fully driverless rides in the quarter, citing regulatory adjustments in some cities.

Baidu’s base is still large. It said cumulative public rides passed 23 million by June 2026. It began fully driverless commercial rides in Dubai in July 2026, through its own app and Uber. It has a plan to reach 100 cities by 2027, which is a plan, not a footprint.

Waymo’s edge is Western-market credibility and current volume. More than 500,000 weekly paid rides against Baidu’s Q2 average of under 80,000 is not a close race this quarter.

Baidu’s edge is China density, lower vehicle cost, and faster moves into the Middle East. If its domestic operations recover as Baidu said they were starting to in August, the gap could shrink again within a few quarters.

That makes the global gap larger than it looked in early 2026. Waymo is now about 6–18 months ahead of Baidu globally, with the low end applying if Apollo Go returns to its Q1 pace.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Is Waymo ahead on manufacturing these days?

Waymo is ahead of U.S. robotaxi competitors on fleet industrialization, but this is still an execution risk, not a solved advantage.

The fleet number is now concrete. Waymo had more than 4,000 vehicles in the U.S. by mid-2026, including about 300 Ojai minivans, according to reports around its September launches. Its June 2026 recall covered nearly 3,900 robotaxis, which fits that range. Most of that fleet passes through the Mesa, Arizona integration facility built with Magna, which Waymo said in 2026 was scaling toward capacity of tens of thousands of vehicles a year. That capacity is a goal, not output.

The second signal is vehicle diversification. Jaguar discontinued the I-PACE, so Waymo is moving to the Zeekr-based Ojai, now in public service, and to the Hyundai IONIQ 5, which is in testing with specialists on board.

The third signal is cost-down engineering. Waymo’s sixth-generation Driver uses fewer sensors than earlier generations but higher-performance sensing, including 17-megapixel imagers, lidar, imaging radar, and external audio receivers.

But scaling fleets is more than putting sensors on cars. Waymo still has to scale cleaning, charging, maintenance, depots, spare parts, insurance, remote assistance, and vehicle uptime in 15 cities at once. Doubling rides to its 1 million weekly target would roughly require doubling those operations.

That puts Waymo about 18–30 months ahead of U.S. robotaxi rivals on fleet industrialization. Against Chinese players, the lead is less clear: Pony.ai said its global fleet passed 1,700 vehicles in May 2026 and targets 3,500 by year-end, and WeRide reported about 1,000 vehicles in China by April.

Is Waymo ahead on unit economics these days?

Waymo is probably ahead on U.S. robotaxi unit-economics learning, but public data still does not prove attractive margins.

The strongest positive signal is utilization. A fleet giving more than 500,000 paid weekly rides learns what matters: trips per vehicle, deadhead miles, charging downtime, cleaning cycles, remote assistance frequency, pickup friction, and city-by-city demand. Uber’s CEO said in 2025 that Waymo vehicles on Uber in Austin were busier than 99% of human drivers on the platform.

The second signal is capital access. In February 2026, Waymo announced a $16 billion investment round at a $126 billion post-money valuation. In October 2026, it closed a $5 billion term loan, its first debt financing, led by PIMCO, Blackstone, and Sixth Street. Lenders are more cautious than equity investors, so the loan suggests the fleet is starting to look like an asset that can carry debt. It still does not prove profitability.

The third signal is cost-down architecture. The sixth-generation Driver, the Ojai vehicle, the Hyundai partnership, and the Mesa facility all point toward a lower cost per deployed vehicle.

But we should not overstate the evidence. Waymo does not disclose fully loaded cost per mile, contribution margin per trip, remote assistance costs, depot labor, insurance, depreciation, or city-level profitability. Research on California data also suggests ride-hailing AVs can generate deadhead miles similar to human-driven ride-hailing.

So Waymo is ahead on the learning curve, not on proven profitability. It is probably 12–24 months ahead of U.S. rivals on unit-economics evidence because nobody else has comparable paid utilization.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Is Waymo ahead on distribution today?

Waymo is ahead in U.S. robotaxi distribution because it can use both its own app and Uber.

That two-channel model matters. A robotaxi company needs safe autonomous supply and rider demand. Waymo’s own app gives it a direct customer relationship, now with features like teen accounts, launched in Nashville in September 2026 for riders aged 13 to 17 linked to a parent. Uber gives it access to a large demand pool in Austin and Atlanta.

The app channel is also growing on its own. Waymo said more than 100,000 people joined from the Houston interest list before it opened Houston to all riders in August 2026. Partnerships add demand at key points, such as the August 2026 deal making Waymo the official ride-hailing partner of the Las Vegas Raiders and Allegiant Stadium.

Competitors are moving toward similar models. Baidu sells rides through Uber in Dubai and tests with Uber and Lyft in London. Wayve works with Uber in London. Pony.ai and WeRide use ride-hailing and government-backed channels. The market is converging on AV companies supplying cars and platforms supplying riders.

Waymo’s edge is that it already runs both routes at scale: its own app in most cities and Uber in Austin and Atlanta.

That makes Waymo about 12–24 months ahead of U.S. competitors on channel flexibility. Globally, Baidu is close because it is building international platform partnerships.

Is Waymo still ahead with regulators today?

Waymo is ahead in U.S. regulatory maturity, but regulators no longer treat it like a protected experiment.

The positive case is clear. Waymo holds driverless deployment permissions, runs paid service in several states, has navigated California CPUC and DMV processes, and has years of work with public agencies. It is not asking whether robotaxis should exist but negotiating how fast they expand.

Competitors face steeper questions. Zoox’s federal exemption comes with a fleet cap and enhanced oversight, and it still needs California deployment permits. Tesla’s rollout faces NHTSA questions and years of controversy around Full Self-Driving claims. In China, the Wuhan outage showed how fast a regulator can freeze a whole sector’s growth.

Waymo’s own scrutiny has increased. School-bus behavior, the Santa Monica child collision investigation, the San Antonio floodwater recall, and the freeway construction-zone recall show that Waymo is now regulated like public infrastructure. None of the open NTSB or NHTSA cases has produced a public finding yet.

So the regulatory lead is real, but it comes with a burden. Smaller companies face fewer investigations because they have fewer cars in fewer places. Waymo gets more scrutiny because it is the one operating at scale.

That puts Waymo roughly 18–36 months ahead of U.S. rivals on regulatory operating maturity.

Is Waymo ahead on public trust right now?

Waymo is ahead on public trust in the U.S., but that trust is still fragile.

The best trust signal is usage. Waymo says it has completed more than 20 million trips and gives more than 500,000 paid rides a week. People do not keep using robotaxis at that scale if the experience feels scary or unreliable. In October 2026, Waymo said San Antonio, San Diego, and Tampa had welcomed hundreds of thousands of riders since opening.

The second signal is rider satisfaction. Uber said Waymo rides in Austin averaged 4.9 stars, a practical signal because riders rate the full trip: pickup, waiting, driving style, comfort, and drop-off.

The third signal is product design and access. The Ojai cabin is built around autonomy, with a flat floor, large screens, easier entry, braille, and screen-reader support. Letting parents send teens alone in Nashville is a trust test of its own.

But trust can fall faster than it rises. Videos of school-bus issues, flooded-road behavior, or collisions involving children shape opinion more than millions of quiet trips. The Wuhan stall showed how one fleet-wide failure can damage a leader’s image overnight.

So Waymo is probably 12–24 months ahead of U.S. competitors on public trust because it has the most real rider exposure. The lead is fragile because one high-profile failure can erase months of quiet gains.

Is Waymo’s data moat still real today?

Yes. Waymo has the strongest deployed robotaxi data moat, but Tesla and Baidu challenge it from different directions.

Waymo’s strongest data is not raw miles. It is driverless ride-hailing data: paid trips, pickups, drop-offs, remote assistance events, rider behavior, urban edge cases, emergency interactions, depot operations, and city-by-city feedback.

The scale is meaningful. Waymo reports more than 270 million rider-only miles through June 2026 and more than 20 million completed trips. That dataset is hard to recreate because a company needs vehicles, riders, regulators, driverless permission, and city diversity at once.

Tesla’s counterargument is fleet scale. It has millions of consumer cars collecting supervised driving data. If that data turns into safe unsupervised behavior, Tesla’s learning loop could become larger than Waymo’s. But today its unsupervised robotaxi mileage is in the hundreds of thousands, not hundreds of millions.

Baidu’s counterargument is China density. Apollo Go passed 23 million cumulative rides by June 2026, with deep operations in Wuhan and Shenzhen. That is a strong dataset, though less transparent to Western observers, and its ride volume fell sharply in Q2.

So Waymo is probably 24–36 months ahead of U.S. rivals on deployed robotaxi data. Versus Tesla, Waymo is ahead on robotaxi-relevant data but not raw driving data. Versus Baidu, the comparison is close on totals, with Waymo ahead on current pace.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Is Waymo ahead on the core autonomy stack today?

Waymo is ahead on validated Level 4 autonomy, but it is not guaranteed to have the most scalable long-term architecture.

Waymo’s stack is built around redundancy. Its sixth-generation Driver combines cameras, lidar, imaging radar, external audio receivers, custom silicon, cleaning systems, and sensor fusion. The logic is simple: a robotaxi that must beat human safety should not depend on one sensor type in darkness, rain, or around emergency vehicles and pedestrians.

That approach has the strongest deployed validation today. Waymo has far more public evidence of paid, rider-only Level 4 driving in major cities than Tesla, Zoox, or Wayve.

But the long-term question remains open. Tesla’s camera-first approach could be cheaper to scale if it proves safe, and Musk says FSD v15 is the release meant to unlock that. Wayve’s end-to-end approach could generalize better across cities. Baidu benefits from low-cost China manufacturing, though the Wuhan stall raised questions about fleet-level software reliability.

So the current answer is split. Waymo is ahead in validated Level 4 autonomy today. It is not guaranteed to win the final cost curve.

That makes Waymo roughly 24–48 months ahead in proven urban Level 4 technology, with a real chance that a simpler architecture compresses the gap later.

Is Waymo ahead internationally these days?

Not yet. Waymo is the U.S. leader now exporting its model, while Chinese players already sell rides abroad.

Waymo’s international plan is now concrete. It has been testing in London since April 2026 and still says it aims to launch rides there by the end of 2026, subject to government approval. That date is not secured. In August 2026, Transport for London said no operator had yet applied for the permit a driverless service needs, and the UK government talks of 2027 for the full framework. Uber and Wayve beat Waymo to London in September 2026, with a safety driver on board. In August 2026, Waymo announced Munich for 2027. In October 2026, it announced Toronto, aiming for first riders in 2028, and the Madrid region, aiming for first riders in 2027. It has also done groundwork in Tokyo. Its October 2026 debt financing was explicitly raised to fund U.S. and international expansion.

Baidu is further along abroad. Apollo Go started fully driverless commercial rides in Dubai in July 2026, began open-road testing in London with Uber and Lyft the same month, received fully driverless testing permits in Hong Kong, and tests in Switzerland with PostBus. WeRide and Pony.ai are active in the Middle East and Asia. Wayve is well placed in the UK through Uber.

Waymo’s advantage abroad is trust transfer: its U.S. safety record, Alphabet backing, and regulatory maturity carry weight in London, Madrid, and Tokyo. Its weakness is localization speed. Each market brings new roads, rules, politics, labor concerns, and vehicle regulations.

So Waymo is not the global rollout leader today. Chinese AV companies are about 6–18 months ahead in Middle Eastern and China-linked markets, while Waymo’s first paid rides outside the U.S. are still ahead of it.

So finally, how far ahead is Waymo today?

Waymo is roughly 24–36 months ahead of the U.S. robotaxi field and about 6–18 months ahead of Baidu globally.

That estimate comes from the full operating stack. Waymo says it gives more than 500,000 paid rides a week in 15 U.S. cities, with a fleet reported at more than 4,000 vehicles. It has more than 270 million rider-only miles, the strongest public safety evidence, a two-channel distribution model, a scaling manufacturing path, $21 billion of new equity and debt raised in 2026, and years of regulator experience.

Against Tesla, the gap is still around 30–48 months: its Texas fleet and Nevada permit are real progress, but it has not shown large-scale unsupervised paid service. Against Zoox, now charging fares in Las Vegas, the commercial gap is around 24–36 months. Against Cruise, the gap is permanent because Cruise left the race. Against Baidu, the gap widened in 2026 because the Wuhan outage cut Apollo Go’s Q2 rides to about a third of its Q1 level; if Baidu recovers fully, the gap could narrow again. Abroad, Baidu still leads.

The most important point is that Waymo’s lead is compound. It is ahead not just on safety or rides or cities, but because those layers reinforce each other. More rides produce more data. More cities improve operations. More safety evidence helps regulators. More regulators enable more cities. More fleet learning improves economics.

If you want more recent data on this point, please see our latest autonomous vehicle market report.