Pony.ai is targeting 3,000 robotaxis across more than 20 cities globally by the end of 2026 — nearly half of them outside China. The company reported its first quarterly profit on Thursday: $75.5 million. The number looked like vindication. It was not. The windfall came from an early investment gain, not from ferrying passengers. The robotaxi business itself is still losing money. What Pony.ai actually announced this week was something more interesting than an accounting quirk: a Chinese autonomous vehicle company with real operational data from dense urban cities is now putting that playbook into markets outside its home turf, through partnerships with local transport operators. Whether it is the first to do so at commercial scale is a claim the evidence does not yet settle cleanly — what is clear is that the overseas push is real, and it is accelerating.
The company's fleet has grown from under 300 robotaxis to 1,446 in roughly a year, according to its earnings release. That overseas ramp is the number to watch, and it is already underway.
In Zagreb, Croatia — yes, Zagreb — Pony.ai partnered with Uber and a local operator called Verne to launch what they describe as Europe's first commercial robotaxi service. Initial deployment work is already underway, including public-road validation. In Doha, Qatar, fare-charging rides are live through a partnership with Mowasalat Karwa, the state-owned transport operator. In Dubai, fully driverless approval is expected later this month. Additional cities — Seoul, Hong Kong, Singapore, Luxembourg — are listed on the roadmap, according to a company statement and reporting by Fortune.
"We are unlikely to run large-scale commercial operations in the U.S. anytime soon," CEO James Peng told Fortune. The reason is not technological. Data security concerns have effectively carved Pony.ai out of the American market. So the company is building elsewhere, in markets where Western AV companies face their own friction, and where Chinese AV firms are finding willing partners.
The financial picture is more complicated than the headline profit suggests. Pony.ai raised $413 million on the Nasdaq in late 2024 and an additional $863 million through a secondary listing in Hong Kong one year later — a rare dual-path to public capital. It ended 2025 with $1.5 billion in cash and equivalents. But it also reported a net loss of $76.8 million for the full year, after burning $152.2 million in the first nine months alone on $156.9 million in R&D, according to its SEC filings. The stock is down roughly 30 percent from its Hong Kong IPO price.
The company claims unit economics breakeven in Shenzhen as of February, following Guangzhou in late 2025 — a milestone worth tracking, but a city-level result is not a business-model proof. Pony.ai says its Gen-7 robotaxis in Shenzhen reached 394 yuan (about $54) in daily revenue per vehicle on March 22, completing 25 orders per car. Those are real numbers from a real operating environment, and they are better than anything the company was posting a year ago. But the broader fleet is still subsidized by capital markets, not by rides.
The global robotaxi map is being drawn by geopolitics as much as by technology. Waymo has partnerships in London and Tokyo. The United States remains the primary operating ground for American robotaxi companies — while Pony.ai, WeRide, Baidu Apollo and their peers are building in parallel in markets that Western AV firms have not yet reached at scale, with vehicles, regulatory frameworks, and data governance structures built around Chinese technical standards.
This is not a story about the technology being ready or not ready. The technology is advancing rapidly. The Gen-7 numbers out of Shenzhen show a vehicle operating at a revenue level that would have sounded optimistic twelve months ago. The question is where that technology lands, and who controls the infrastructure it runs on.
Somewhere in a Pony.ai vehicle in Beijing or Shenzhen, a passenger has fallen asleep in the back seat. The car, apparently, has figured out what to do about it. These edge cases — the sleeping riders, the vehicles that chirp at strangers to close doors, the cleaning crews who handle what passengers leave behind — are not bugs in the system. They are the product working. They are also the new workforce story hiding inside the autonomous vehicle headline: the human backstop that makes the machine look seamless.
Pony.ai calls its growth plan a "dual engine" strategy — autonomous driving plus licensing. That framing serves investor decks. The more honest summary: a Chinese company with real operational data from dense urban Chinese cities is deploying that playbook into markets where Western competitors cannot easily follow, using partnerships with local transport operators as its beachhead. Fare-charging is live in Doha. Validation is underway in Zagreb. Dubai is waiting on regulators. The overseas expansion is not a promise anymore — it is already well underway. Whether those markets stay open, whether the economics hold, and whether the company can keep raising capital while posting losses — those are the open questions.
