Ask AI · How does the co-constructed fleet model support Xiaoma Zhixing’s aggressive expansion goals?

01. Commercialization Reflects Reality

On March 26, Xiaoma Zhixing released its first annual report after listing in Hong Kong stocks. The report shows that by 2025, Xiaoma Zhixing’s total revenue will reach $90 million (about 629 million RMB), a 20% year-over-year increase. Among them, the company’s autonomous driving mobility services, autonomous driving truck services, and technology licensing and application businesses all achieved year-over-year growth, especially autonomous driving mobility services, with revenue increasing by 128.6%.

Meanwhile, Xiaoma Zhixing also achieved its first quarterly profit. The report indicates that in Q4 2025, the company’s net profit was $75.45 million (about 528 million RMB), whereas in the same period last year, it posted a net loss of $181 million.

However, the capital market does not seem to buy into Xiaoma Zhixing’s historic breakthrough. On March 27, the first trading day after the earnings release, Xiaoma Zhixing’s Hong Kong stock price opened sharply lower, closing down nearly 14%. On March 31 and April 2, Xiaoma Zhixing’s stock price was further hammered, with the closing price hitting a new low since listing.

Image / Photo by “Financial World”

A large part of Xiaoma Zhixing’s stock price decline is due to the fact that the profit in Q4 2025 was somewhat incidental. Earlier, Xiaoma Zhixing invested in Moore Threads. After Moore Threads went public on the A-share market and its stock price soared, Xiaoma Zhixing, as a shareholder, saw the value of its holdings also rise sharply. In Q4 2025, Xiaoma Zhixing’s “trading financial assets fair value change gains” reached $132 million, widely considered as the unrealized gains from Moore Threads’ listing.

Excluding such effects, Xiaoma Zhixing’s non-GAAP net loss in the same period was $49.04M (about 343 million RMB), an 18.7% increase compared to the same period last year; for the full year 2025, the company’s net loss was $174 million (about 1.22B RMB), a 31.5% increase year-over-year. In other words, Xiaoma Zhixing’s situation of increasing revenue but not profit has raised market concerns.

Compared to the ongoing operating expenses under current business expansion, the biggest highlight of Xiaoma Zhixing’s 2025 financial report is that it proves Robotaxi can be profitable once it reaches a certain milestone.

Today, autonomous driving mobility services have become Xiaoma Zhixing’s core growth engine. In Q4 2025, revenue from this segment grew 160% year-over-year, with passenger fare revenue, which better reflects user willingness to pay, increasing by over 500%. With the seventh-generation Robotaxi officially commercialized, Xiaoma Zhixing has achieved break-even (UE) for single vehicles in Guangzhou and Shenzhen.

02. After Making Money per Vehicle, More Aggressive Expansion Goals

Xiaoma Zhixing co-founder and CFO Wang Haojun provided a detailed breakdown of the economic model of Robotaxi vehicles.

Wang Haojun explained that the costs of Robotaxi roughly fall into two parts: hardware depreciation and operating costs, each accounting for about 50% of total costs.

Previously, the lack of scale effects in Robotaxi was mainly due to high per-vehicle costs, meaning the more vehicles deployed, the greater the losses. Until the seventh-generation Robotaxi, Xiaoma Zhixing made many optimizations in vehicle costs. Compared to the previous generation, the BOM (Bill of Materials) cost of its ADK (Autonomous Driving Kit) decreased by 70%. This year, these costs will continue to be optimized, with a further 20% reduction expected based on 2025 levels.

Image / Photo by “Financial World”

The large-scale deployment of the seventh-generation Robotaxi is key to Xiaoma Zhixing achieving break-even (UE) in Guangzhou and Shenzhen. By the end of 2025, Xiaoma Zhixing’s entire Robotaxi fleet will have 1,149 vehicles, and by the release of the 2026 Q1 financial report, this number will be 1,446. Most of the nearly 300 new seventh-generation Robotaxis during this period are deployed in Shenzhen.

“We saw explosive growth in Shenzhen’s order volume this year. By mid-February, Shenzhen’s commercial order volume had already exceeded the total for all of 2025,” Wang Haojun said.

In the company’s view, the surge in Shenzhen’s commercial orders this year not only proves the replicability of the vehicle profitability model but also indicates that the fleet density increase is creating a positive network effect. Therefore, the company has set a more aggressive target for 2026: deploying in over 20 cities domestically and internationally, with a fleet size of over 3,000 vehicles, and more than tripling Robotaxi revenue.

But to achieve this phased goal, Xiaoma Zhixing needs to increase capital expenditure by tens of thousands of times beyond the 3 billion RMB figure. “No company can expand entirely on its own,” Wang Haojun said.

Thus, Xiaoma Zhixing proposed the “co-built fleet model.” Under this model, partners are responsible for funding vehicle purchases, while Xiaoma Zhixing remains responsible for AI drivers. If partners utilize the company’s customer acquisition platform and fleet operation management, they can also share in the revenue. In this light-asset model, the company can rapidly expand its fleet without large amounts of its own capital. Previously, Wang Haojun revealed at the Q4 earnings call that nearly half of the new vehicles this year would be deployed through this model.

03. Using Technological Depth and Operational Efficiency to Compete

Besides the commercial viability of Robotaxi and the pace of fleet expansion, the external market competition landscape is also a concern. Tesla, Xpeng, Volkswagen, Mercedes-Benz, and others have all entered the Robotaxi field, and Didi’s autonomous driving project has also entered commercial testing this year. As a startup, how does Xiaoma Zhixing stand out in the crowded Robotaxi arena dominated by giants?

Xiaoma Zhixing CTO Lou Tiancheng stated at the above earnings meeting that Robotaxi is an extremely complex systems engineering project, where technology, policy, mass production, operations, and ecological cooperation are interconnected. It’s not simply a matter of resource input to achieve rapid development.

To ensure safety significantly better than human drivers, the system must continuously evolve through large-scale trial and error in virtual environments, with a world model being indispensable. “That’s why L4 autonomous driving requires years of AI investment accumulation, rather than just collecting more real-world data,” Lou Tiancheng said.

He explained that the hardest part of L4 is not the first 99%, but the last 1%—those rare but safety-critical tail scenarios. Currently, only world models can fully cover tail scenarios, and only data from fully autonomous Robotaxis can continuously narrow the gap between the world model and reality.

With technological support, Xiaoma Zhixing’s operational advantages will be amplified. Wang Haojun said that operating costs are an important factor. Although automakers and other platforms have reusable experience in vehicle manufacturing and operation, Xiaoma Zhixing also has its own unique advantages.

“Many companies’ Robotaxi vehicles have not yet been produced, and their operations are difficult to optimize. Whoever can operate faster and achieve larger-scale autonomous driving operations will face challenges, but we are very confident,” Wang Haojun stated.