（SINGAPORE, 17.11.2025）Experts predict that by the end of this year, there will be around 5,000 driverless Robotaxis operating on China’s roads, a figure expected to surge to 632,000 by 2030. Yet Chinese media note that investors remain cautious, as there is no clear evidence that Robotaxis will meaningfully replace traditional taxis in the near term.
On November 6, two Chinese autonomous driving firms operating Robotaxi services—WeRide (文远知行) and Pony.ai (小马智行)—saw their Hong Kong IPOs debut below issue prices. Fundraising was successful, but market sentiment was muted.
“Back in 2021, investors were very enthusiastic about autonomous-driving taxis; the sector was seen as highly promising,” a mobility-sector investor told tech platform Tech Planet (Tech 星球).
However, he noted that as companies expand Robotaxi services and move closer to full-scale operations, a variety of challenges have surfaced—including unclear business models, high operating costs, regulatory hurdles, safety concerns, and weaker-than-expected investor confidence.
“Real-world factors, such as speed in policy implementation and the degree of safety-technology buffers, have become key benchmarks for valuing these firms,” he added.
A central question in commercializing Robotaxis remains: as the technology matures and investment flows in, when will these vehicles actually replace human drivers?
Short-term replacement appears unlikely, though Robotaxis are expected to gradually take on a larger role over the long term. From a user perspective, the service still lags behind ride-hailing and traditional taxis in convenience.
A Shanghai passenger told Tech Planet that using a Robotaxi requires going to a designated pick-up point: “Driverless taxis operate only in a limited area with fixed boarding and drop-off locations, unlike ride-hailing services that offer true door-to-door convenience.”
The fixed pick-up and drop-off system has advantages in certain scenarios. Many points are near bus or subway stations, making transfers easy for commuters. Yet for everyday ride-hailing needs—like saving time or reducing walking—these constraints may deter potential users.
Another current issue is the limited number of Robotaxis. A Guangzhou user reported taking the same driverless taxi multiple times, saying, “There are just too few Robotaxis on the road, and it’s often hard to get a ride when you need one.”
Today, top Chinese companies such as Pony.ai, WeRide, and Luobo Kuaipao (萝卜快跑) have each deployed nearly 1,000 autonomous taxis across domestic and international markets, covering business commuting, short-distance trips, and shuttle services. Data shows Luobo Kuaipao has driven over 100 million km safely, with WeRide at roughly 55 million km and Pony.ai at about 48 million km.
Meanwhile, Qianli Technology （千里科技）, Hello （哈啰）, and DiDi Autonomous Driving （滴滴自动驾驶）are accelerating their Robotaxi initiatives. China’s Robotaxi market is projected to grow from US$54 million (≈S$70 million) this year to US$1.2 billion by 2030, reaching US$4.7 billion by 2035. Between 2025 and 2035, the market is expected to expand an astonishing 757-fold, positioning Robotaxis as a transformative element of future mobility.
Despite the growth potential, industry experts remain discreet. According to one professional, over the next three to five years, adoption is likely to concentrate in specific settings such as business parks, airports, and university campuses. Achieving full citywide competition with ride-hailing and traditional taxis is still a distant prospect.
Profitability remains a challenge. US Waymo co-CEO Dmitri Dolgov revealed that upgrading a fifth-generation Robotaxi with specialized hardware costs about US$100,000, bringing the total per-vehicle cost, including the base car, to US$175,000.
Chinese Robotaxi manufacturers have reduced costs, but prices remain high: Baidu’s Apollo sells for 205,000 yuan (≈S$37,500), while Pony.ai’s seventh-generation model costs about 300,000 yuan—both well above conventional taxi acquisition costs.
Operating costs are also substantial. A Goldman Sachs report shows vehicle depreciation accounts for around 43%, shared fleet operational expenses about 44%, and remote safety operator costs 13%.
Under these conditions, in 2024 Pony.ai reported a net loss of 1.967 billion yuan, while WeRide’s loss was 2.517 billion yuan. In first half of 2025, Pony.ai’s loss widened 75.1% year-on-year to 681 million yuan, whereas WeRide’s narrowed 10.3% to 792 million yuan. Globally, no Robotaxi company has achieved large-scale profitability.
Current pricing provides little competitive advantage. In Shanghai, Pony.ai charges 14 yuan for the first 3 km, with distances beyond 10 km billed at 30% above the standard per-km rate. This approximately matches ride-hailing fares but is insufficient to cover costs, let alone generate profit.
Goldman Sachs projects that Robotaxis in first-tier Chinese cities could achieve positive gross profit per vehicle by 2026, with full operating profitability expected around 2032. Second-tier and smaller cities will likely see slower progress. Pony.ai CEO James Peng predicts equilibrium point around 2028 if its fleet surpasses 50,000 vehicles.
An investor emphasized that commercial viability requires first achieving vehicle-level breakeven—covering each car’s operating costs. Only through scaling and spreading fixed costs can a profitable model emerge and support market share growth.
In short, while Robotaxis offer good financial potential, realizing it will require carefully balancing cost management with fleet expansion, Tech Planet concluded.
