Pony.ai reports first profitable quarter and plans robotaxi expansion to over 20 cities

Pony.ai showed quarterly profit for the first time, but earned it not from robotaxi rides themselves, but thanks to income from an early investment. At the same time, the company is sharply raising its targets for 2026: it wants to deploy the service in more than 20 cities and increase the fleet to more than 3000 vehicles.

Where the Profit Came From

In the fourth quarter of 2025, Pony.ai received net profit under GAAP of $75.5 million.

For the company, this is the first such quarter in its history. At first glance, the news looks like a long-awaited turning point for the autonomous transport sector, which has lived on investments and promises of future scalability for many years. But in the details, the picture is more cautious: the main contribution to the overall positive result was not the robotaxi service itself, but income from investments and revaluation of securities.

Pony.ai's annual revenue grew 20% and reached $90 million, and net loss for all of 2025 shrank 72% — to $76.8 million.

At the same time, the company still has operating losses, meaning its core business still requires significant investment. In other words, this is an important financial milestone, but not proof that robotaxis have already become a stable profitable mass service.

"We are making ahead-of-the-curve investments to accelerate commercialization," said

Pony.ai CFO Leo Wang in the company's report.

Plan for 20 Cities

Pony.ai's main focus right now is not on a nice quarterly report, but on scale. The company announced that as of March 25, 2026, it has produced 1446 robotaxis and intends to bring the fleet to more than 3000 vehicles by year end. Operating geography is also expanding rapidly: the goal is presence in more than 20 cities worldwide, with nearly half of them located outside China.

• 1446 robotaxis already produced by the end of March 2026
• Target for 2026 — fleet exceeding 3000 vehicles
• 1000 Toyota bZ4X secured for joint deployment in 2026
• New markets include Hangzhou, Changsha, Zagreb, Doha, and Singapore
• In Dubai, the company is awaiting formal approval for fully autonomous operations

Pony.ai wants to drive such growth through a model of joint deployment with automakers and local partners. The company particularly emphasizes the alliance with Toyota, as well as cooperation with BAIC and GAC. The logic is clear: if the company can rely on ready-made production, service networks, and local operators, then scaling will be faster and require less capital than building everything alone. This is especially important for an industry where equipment costs and fleet maintenance usually eat into growth economics.

Economics of Rides

Despite the importance of investment income, Pony.ai does have signs of progress in the core business. Revenue from the Robotaxi segment in the fourth quarter grew 160% year-over-year — to $6.

7 million. Income from paid rides, according to the company, increased more than 500% for the quarter. For all of 2025, robotaxi revenue reached $16.

6 million, which is 128.6% more than a year earlier. Even more importantly, there's an announcement of unit economics breakeven in Guangzhou and Shenzhen just four months after launching seventh-generation robotaxis.

This does not mean the entire company has become profitable. But it does mean that in certain cities and at the level of individual vehicles, the economics are beginning to converge: ride revenue covers direct operating expenses significantly better than before. For the market, this is one of the few practical signals that the model can scale without a sharp increase in losses per new vehicle.

Pony.ai backs this up with concrete metrics. In Shenzhen on peak day March 22, 2026, net daily revenue per vehicle reached 394 yuan, and the number of orders — 25 per vehicle.

By the end of March, the user base in China approached 1 million people. The company also claims that the number of paid orders in Shenzhen from January to mid-February 2026 already exceeded the level for all of 2025, which appears to be a sign of rapidly growing demand.

What This Means

Pony.ai has not yet proven it has learned to earn money from robotaxis at company scale, but has already shown that individual cities and vehicles can converge economically. If the plan for 20+ cities and a fleet of 3000+ vehicles works out, the market will begin evaluating the company not by one-off investment income, but by recurring revenue from autonomous rides.