Is China winning the robotaxi race?
In our autonomous vehicle market deck, you will find everything you need to understand the market
SUMMARY
China is winning the robotaxi race as a country, but Waymo still runs the strongest individual service.
China’s advantage comes from depth. Baidu, Pony.ai and WeRide are all operating paid driverless services, expanding fleets and working with major vehicle manufacturers; the United States still depends overwhelmingly on Waymo for proven scale.
The ride-volume comparison is less clean than the headlines suggest. Waymo leads disclosed weekly paid rides, while Baidu leads disclosed cumulative public rides, and China probably moves more passengers overall once Pony.ai and WeRide are included.
China already has the larger combined fleet. Its next advantage may be more important: several manufacturers can produce purpose-built robotaxis in the thousands rather than treating each vehicle as a bespoke technology project.
Vehicle cost is where China looks hardest to catch. Baidu and Pony.ai are pushing complete robotaxis toward ordinary-EV prices, while Waymo has not disclosed the finished cost of its newer Ojai platform.
Waymo’s clearest lead is safety evidence. It publishes detailed crash comparisons across a huge rider-only mileage base, and independent IIHS research supports the claim that its vehicles crash less often than human drivers in the same cities.
Nobody has proved the business model. Pony.ai and WeRide are growing robotaxi revenue but still losing tens of millions of dollars per quarter, while fleet-level breakeven claims do not include the full cost of research, software and expansion.
The Wuhan outage exposed a risk that grows with scale: one software or fleet-control failure can stop many cars at once. China’s regulators responded with a permit pause, showing that fast expansion still depends on public confidence.
Chinese companies have built the stronger overseas pipeline, especially in the Middle East and parts of Asia. Yet most foreign launches remain small, and US security rules effectively divide the market into regional races rather than one open global contest.
The sharpest conclusion is that China leads robotaxi industrialization—competition, manufacturing, fleet growth and cost—while Waymo leads current service quality, weekly usage and transparent safety proof. China is ahead overall, but it has not produced a single operator better than Waymo yet.
This market map, featured in our autonomous vehicle market deck, highlights top companies and startups in the autonomous vehicle market
Why does China look like the robotaxi leader right now?
China looks ahead because it has turned robotaxis into a real multi-company industry, while the United States still relies heavily on Waymo.
Baidu’s latest quarterly report says Apollo Go completed 3.2 million fully driverless rides, more than double the previous year’s volume. Pony.ai now has more than 1,700 robotaxis, and WeRide reported about 1,300 worldwide, including roughly 1,000 in China. All three are charging passengers, adding vehicles and opening new operating areas.
No other country has three robotaxi companies at this level. The United States has Waymo at genuine scale, followed by a large gap to Tesla, Zoox and smaller programs.
Who looks ahead depends on what we count. China leads in fleet building, manufacturing and the number of serious operators. Waymo leads in weekly paid rides and public safety proof.
| Measure | Leader now | What we found |
|---|---|---|
| Current weekly ride volume | Waymo | The largest disclosed paid service |
| Cumulative public rides | Baidu Apollo Go | The largest disclosed public total |
| Vehicle cost | China | Purpose-built cars around the price of a normal EV |
| Public safety evidence | Waymo | Detailed results across a large rider-only mileage base |
| Overseas expansion pipeline | China | More companies testing and launching in more countries |
| Overall profitability | Nobody | Commercial revenue remains far below total spending |
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Is China already giving more robotaxi rides than the United States?
China probably handles more driverless taxi trips in total today, but the published numbers are too inconsistent to measure the gap cleanly.
Baidu’s 3.2 million rides in one quarter work out to about 246,000 per week. Its busiest disclosed week passed 350,000. Waymo, meanwhile, is running above 500,000 paid rides per week.
Adding Pony.ai and WeRide probably pushes China above the American total. Pony.ai says its paid orders have been rising every month, while WeRide averaged more than 17 daily orders per vehicle and reached 28 at busy times. Neither company publishes a weekly total that lines up neatly with Baidu’s or Waymo’s definition.
As this chart shows, and as featured in our autonomous vehicle market deck, search interest in autonomous vehicles has continued to rise
Has Baidu Apollo Go overtaken Waymo?
Waymo runs the stronger robotaxi service today, even though Baidu has edged ahead on cumulative rides.
Apollo Go has provided more than 22 million public rides, compared with Waymo’s figure of more than 20 million. Baidu wins the historical count. Current use points the other way, with Waymo carrying roughly twice as many passengers per week as Baidu averaged during its latest reported quarter.
There is one important catch. Waymo calls its number paid rides, while Baidu uses broader terms such as “fully driverless operational rides” and rides provided to the public. We cannot tell how many Baidu trips were discounted or promotional, so the totals are close but not perfectly comparable.
| Metric | Baidu Apollo Go | Waymo |
|---|---|---|
| Cumulative disclosed rides | More than 22 million | More than 20 million |
| Latest weekly level | Peak above 350,000 | More than 500,000 paid rides |
| Public operating footprint | Large Chinese zones plus early overseas services | More than 10 US metro areas |
| Strongest advantage | Low-cost fleet expansion | Current usage and safety proof |
Does China have the world’s biggest robotaxi fleet?
China almost certainly has the world’s biggest combined robotaxi fleet, but no Chinese company matches Waymo’s roughly 4,000 vehicles.
Pony.ai alone has passed 1,700. WeRide reported about 1,300 globally, and the Associated Press said Baidu operates more than 1,000, mostly in China. Those three fleets already add up to at least 4,000 before counting AutoX, Momenta and other smaller programs.
China’s lead could grow quickly. Pony.ai raised its year-end target above 3,500 vehicles. WeRide expects 2,000 new Geely-built GXRs to lift its active fleet beyond 2,600. Not all those vehicles are carrying passengers yet, although signed production plans make the targets more concrete than a distant corporate promise.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
This chart, included in our autonomous vehicle market deck, illustrates yearly VC funding for autonomous vehicle startups
Are China’s robotaxis really fully driverless?
Yes, the leading Chinese robotaxi services are genuinely driverless today, but only inside approved zones and operating rules.
Baidu says its mainland services have run without in-car safety drivers since early 2025. Pony.ai has fully driverless commercial Gen-7 vehicles in Beijing, Guangzhou and Shenzhen. WeRide runs a 24-hour driverless service across a large part of Guangzhou, including airport and rail connections.
The cars may have nobody in the front seat, yet they follow approved maps, roads, hours, weather limits and pickup rules. A robotaxi can drive itself across one district and remain unable to serve the next one.
Remote staff can also help when a car gets confused. Usually they provide context or approve a safe next step instead of steering every movement. Those teams cost money, so removing the in-car driver does not remove every labor expense.
Is China expanding robotaxis across cities faster than Waymo?
China is opening more locations, while Waymo currently gives more rides in fewer places.
Baidu says it is present in 27 cities. WeRide says its vehicles have been tested or operated in more than 40 cities across 12 countries. Pony.ai wants more than 20 cities by year-end. Those counts mix public services with tests, mapping work and small pilots.
Waymo gives a narrower count. It recently covered more than 1,400 square miles across 11 cities where the public could hail rides, then announced Denver, Las Vegas, San Diego and Tampa as its next markets. Passengers can use Waymo across large zones in its active cities; many Chinese overseas locations contain only a few trial cars.
This chart, included in our autonomous vehicle market deck, shows how Waymo is winning in autonomous vehicles
Are Chinese robotaxis much cheaper to build?
China has the clearest cost advantage in robotaxis, and that advantage comes from its car industry as much as from autonomous-driving software.
Baidu originally priced the purpose-built Apollo RT6 at RMB250,000, or about $37,000, and later said the delivered cost had fallen below $30,000. Pony.ai expects the full cost of a Chinese Gen-7 robotaxi, including the autonomous-driving kit, to drop below RMB230,000 by 2027.
Waymo has never said what a complete robotaxi costs. Its older Jaguar I-Pace fleet combines a premium car with a costly sensor and computing package. The new Ojai uses fewer sensors, a purpose-built Zeekr body and final integration in Arizona, so Waymo is clearly pushing costs down. Until it publishes the finished price, a direct comparison is impossible.
China’s supply chain makes those low figures believable. Baidu works with Jiangling Motors, Pony.ai with BAIC, GAC and Toyota, and WeRide with Geely’s Farizon. All three programs use established car factories and suppliers.
| Robotaxi platform | Disclosed cost or capacity | What it tells us |
|---|---|---|
| Baidu Apollo RT6 | RMB250,000 at launch; later reported below $30,000 | China can build an L4 vehicle near mass-market EV prices |
| Pony.ai Gen-7 | Target below RMB230,000 by 2027 | The whole vehicle and driving kit are being costed together |
| WeRide GXR | 2,000 new vehicles planned | Manufacturing partners can add fleet capacity quickly |
| Waymo Ojai | Final cost undisclosed | Lower-cost design, with capacity for tens of thousands per year in Arizona |
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Can China turn cheaper robotaxis into a profitable business?
Cheap vehicles are improving the math, but Chinese robotaxi companies remain far from proving a profitable business.
Pony.ai’s robotaxi revenue reached $8.6 million in its latest quarter, almost five times the previous year. The company lost $53.5 million and used $74.2 million in operating cash. WeRide produced $16.5 million of total revenue across all its activities and lost $56.4 million.
Baidu gives no separate income statement for Apollo Go, although it says Wuhan was approaching fleet-level breakeven. WeRide has made a similar claim for Abu Dhabi after removing the safety driver. Those claims may be true for the direct cost of running those fleets, but they leave out some company-wide research and expansion spending.
This chart, included in our autonomous vehicle market deck, illustrates yearly funding for autonomous vehicle startups
Are low robotaxi fares proving demand or buying rides?
Chinese riders are using robotaxis more often, but low fares and promotions are doing part of the selling.
Apollo Go has offered base fares as low as RMB4 on routes where a normal taxi might start near RMB18. That gap can attract riders who care more about price than autonomy. A study of 336 real Apollo Go users in Wuhan also found cost sensitivity among the strongest predictors of whether people chose the service.
Paid use is rising too. Pony.ai’s average weekly paid orders rose 119% between January and May, while its registered Chinese user base more than tripled. WeRide says daily orders per vehicle have climbed as its waiting times improved.
The harder test comes later. Demand looks durable when riders keep coming after discounts shrink and the fare approaches the real cost of the trip. China has shown that people will ride; it has only partly shown what they are willing to pay.
Who has the stronger robotaxi safety evidence, China or Waymo?
Waymo currently has the stronger public safety case by a wide margin.
Its latest analysis covers 220.6 million miles with no human driver. Compared with human drivers on the same roads, Waymo reported 94% fewer crashes causing serious or fatal injuries, 82% fewer airbag-deployment crashes and 82% fewer crashes involving any reported injury. The company publishes its method, local benchmarks and regular updates tied to US reporting rules.
A newly released independent study from the Insurance Institute for Highway Safety found that Waymo’s driverless vehicles had 68% fewer police-reported crashes per mile than human drivers in the same cities. That outside result makes the safety lead much harder to dismiss as company marketing.
Baidu has also accumulated a huge amount of experience, including more than 190 million fully driverless kilometers by its previous annual update. Baidu does not release the same kind of crash table, with injuries and serious crashes per mile compared against drivers on the same roads. Pony.ai and WeRide reveal even less.
Waymo’s record includes mistakes. Lately, it recalled software on 3,871 vehicles after cars entered closed freeway work zones, and another update tightened behavior around flooded roads. Those recalls show that Waymo can miss difficult situations. They also let outsiders see the failure, the fleet size and the fix.
| Safety question | Waymo | Leading Chinese operators |
|---|---|---|
| Large driverless exposure base | 220.6 million rider-only miles | Baidu has reported more than 190 million fully driverless kilometers |
| Detailed crash rates | Publicly available | No comparable public breakdown |
| Human-driver comparison | Adjusted for the same operating areas | Not published in matching form |
| Public recall trail | Detailed through NHTSA | Much less visible outside China |
| What we can say | Strong evidence of lower crash rates | Too little public data for the same confidence |
If you want more recent data on this point, please see our latest autonomous vehicle market report.
This chart, included in our autonomous vehicle market deck, compares the main business model options for autonomous trucking companies
Did the Wuhan outage expose a serious weakness in China’s robotaxi model?
The Wuhan outage exposed a serious fleet-level weakness that ordinary taxis do not share.
More than 100 Apollo Go cars stopped during a system malfunction, according to Wuhan police and the Associated Press. Some passengers were stranded in fast-moving traffic, although no injuries were reported. The size of the failure made it different from one car getting confused at an intersection.
One bad engine disables one taxi. A failure in shared software, cloud services or fleet control can affect many robotaxis at once. The precise technical cause in Wuhan remains unknown, so blaming the cloud would go beyond the evidence. A simultaneous failure across the fleet already shows the risk.
The episode tells us more about reliability than crash avoidance. Waymo has faced its own fleet-wide problems, including cars stopping during a power outage and large software recalls. Any city that depends on robotaxis will need backup communications, faster passenger rescue and a way to clear many disabled vehicles at once.
Is China’s regulation helping robotaxis move faster?
China’s regulators continue to help robotaxis move faster, though the recent permit freeze showed that support has limits.
Cities such as Beijing, Guangzhou, Shenzhen and Wuhan have created large testing zones and permits that gradually let companies remove the safety driver. Local governments also want the factories, data centers and technical jobs that come with the industry. That competition has given companies many places to test real services.
After the Wuhan shutdown, authorities paused new autonomous-driving permits and reviewed the sector. Lately, approvals have started to resume in selected cities, with Momenta among the first companies receiving a new permit and Apollo Go showing signs of returning in Wuhan.
China’s approach is less permissive than it sometimes looks. Officials let companies expand quickly, then impose a broad pause when a failure threatens public confidence.
This chart, featured in our autonomous vehicle market deck, shows the share of revenue generated by each customer segment in the autonomous vehicle market
Can Chinese robotaxi companies win outside China?
Chinese robotaxi companies are ahead in overseas deal-making, but only a few foreign markets now host real driverless commercial service.
WeRide has the strongest proof so far. It operates fully driverless rides in Abu Dhabi through Uber, began public operations with Grab in Singapore and has announced a Zurich launch with Uber. It also recently unveiled a right-hand-drive robotaxi with Geely for markets such as the United Kingdom, Singapore and Hong Kong.
Baidu is running fully driverless operations in parts of Dubai and a service with AutoGo in Abu Dhabi. London tests with Uber and Lyft and open-road work in Switzerland are next. Pony.ai says it has entered nine countries and started public services in Croatia, Qatar, Singapore and South Korea.
The headline counts make these launches sound more mature than many of them are. “Operating in a country” can mean several test cars with safety staff. Abu Dhabi is the clearest example of a Chinese company running a genuine foreign driverless service. Most other markets have yet to turn permits and partnerships into regular passenger volume.
Is the United States closed to Chinese robotaxi companies?
The American market is effectively closed to Chinese robotaxi software, which makes a single global winner unlikely.
The US Commerce Department’s connected-vehicle rule restricts autonomous-driving software and connected systems linked to China or Russia. The software restrictions begin with model-year 2027 vehicles, and Chinese-controlled manufacturers face wider limits on selling connected cars in the country.
Waymo’s Ojai shows how narrow the permitted route has become. Zeekr builds the vehicle body in China, then Waymo installs its own sensors, computers and autonomous-driving system in Arizona. Waymo says the imported shell contains no Chinese telematics.
Chinese companies can lead at home and win contracts in the Middle East, Asia or Europe without ever entering the largest American cities. Waymo can dominate the United States while remaining absent from mainland China. Geography and security policy are already splitting the race into regional markets.
This chart, included in our autonomous vehicle market deck, shows how robotaxi platform technology has evolved over time
So, is China winning the robotaxi race?
Partly: China is currently winning the race to industrialize robotaxis, while Waymo leads the race to run the best robotaxi service.
China has the broader field. Baidu, Pony.ai and WeRide are all putting driverless cars into paid service, ordering thousands more, lowering vehicle costs and signing overseas partners. The country’s EV factories and component suppliers give those operators a path from several thousand vehicles to much larger fleets.
Waymo is the better operator today. As seen above, it carries more passengers each week than any Chinese rival, operates the largest one-company fleet and backs its safety claims with a level of public detail that China’s companies have not matched. Its new vehicle and faster city rollout also reduce two of China’s clearest advantages.
Tesla could change the American side of the race later, but it is nowhere near the lead at present. Its latest update showed more than 380,000 unsupervised robotaxi miles across six cities; Waymo now drives more than 4 million fully autonomous miles in one week. Tesla’s planned Cybercab production capacity is worth watching, while its current service remains small.
China still has three big problems: transparent safety data, full-company profitability and the fleet reliability questions discussed above. Waymo has to show that its service can scale internationally and get much closer to Chinese vehicle costs.
China is ahead as a country and as a manufacturing system; Waymo remains ahead as a company. Calling China the outright winner goes too far, but calling the race even misses how much industrial ground China has already taken.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
OUR METHODOLOGY
This analysis tests whether China is winning the robotaxi race by separating several contests that are often bundled together: current ride volume, fleet scale, vehicle economics, safety evidence, international expansion and the path to profitability.
We prioritized activity already happening over long-range promises. Completed rides carried more weight than announced launches, active vehicles mattered more than theoretical factory capacity, disclosed financial results mattered more than breakeven ambitions, and public crash data mattered more than broad claims of safety.
The companies do not report identical metrics. We kept the distinctions between paid rides, public rides, operational rides, active fleets, produced vehicles, pilot programs and commercial services rather than forcing them into one false comparison.
Where the figures were reasonably comparable, we used them directly. Where definitions differed, we treated the evidence as directional. That is why the article can say Waymo leads weekly paid rides while Baidu leads cumulative disclosed public rides, without pretending those two totals measure exactly the same thing.
Financial claims were handled in the same way. Fleet-level breakeven can show that a local service is covering some direct operating costs, but it is not treated as proof that the wider robotaxi business is profitable after research, software development, corporate overhead and expansion spending.
Safety received separate treatment because public evidence varies sharply by operator. Waymo’s rider-only mileage, crash-rate methodology, human-driver comparisons and recall trail allow a stronger conclusion than the more limited public reporting from Baidu, Pony.ai and WeRide.
We did not use a mechanical score to declare a winner. The conclusion comes from the combined direction of the evidence: China has the broader industrial system and lower-cost manufacturing path, while Waymo has the strongest single service, the highest disclosed weekly paid usage and the most transparent safety record.
Key sources include Baidu’s Q1 2026 results, Pony.ai’s Q1 2026 results, WeRide’s Q1 2026 results, Waymo’s latest safety update, the Insurance Institute for Highway Safety’s independent crash comparison, the Associated Press report on the Wuhan fleet malfunction, and the US Bureau of Industry and Security’s connected-vehicle rules.
In our autonomous vehicle market deck, we identify pain points entrepreneurs should prioritize
Related blog posts
