--- title: "Pony.ai's profit alert: Robotaxi, can we really start counting the money now?" type: "Topics" locale: "en" url: "https://longbridge.com/en/topics/38425364.md" description: "Recently, the autonomous driving star company Pony.ai released its first annual forecast after going public: it is expected that the group's net loss in 2025 will narrow significantly from $275 million in 2024 to between $69 million and $86 million. Compared to the full-year loss of $275 million in 2024, Pony.ai's loss in 2025 has shown a significant narrowing. More importantly, this is not simply due to cost compression or one-time subsidies, but rather it is happening at a time when the inflection point of Robotaxi commercialization is being verified..." datetime: "2026-02-04T06:29:30.000Z" locales: - [en](https://longbridge.com/en/topics/38425364.md) - [zh-CN](https://longbridge.com/zh-CN/topics/38425364.md) - [zh-HK](https://longbridge.com/zh-HK/topics/38425364.md) author: "[港股研究社](https://longbridge.com/en/profiles/3199113.md)" --- > Supported Languages: [简体中文](https://longbridge.com/zh-CN/topics/38425364.md) | [繁體中文](https://longbridge.com/zh-HK/topics/38425364.md) # Pony.ai's profit alert: Robotaxi, can we really start counting the money now? Recently, the autonomous driving star company Pony.ai released its first annual forecast after going public: it is expected that the group's net loss in 2025 will narrow significantly from $275 million in 2024 to $69 million to $86 million. Compared to the full-year loss of $275 million in 2024, Pony.ai's loss in 2025 has narrowed significantly. More importantly, this is not simply due to cost compression or one-time subsidies but occurs at a time when the inflection point of Robotaxi commercialization is being validated. While the entire Robo-X track is still in a stage where demonstration outweighs commercialization, Pony.ai's positive trading update seems to be a signal that L4 autonomous driving can begin to move out of the lab and into the financial modeling phase. **Not Just a Positive Update, but a Business Model Taking Shape** According to the company's announcement, the direct reason for the narrowing of Pony.ai's losses in 2025 is the increase in fair value gains from investments in listed companies. In other words, its short-term positive performance is not entirely contributed by its core business. But this is not the most important; what truly deserves attention is the "structural change" happening simultaneously in its core business. As early as the third quarter of 2025, Pony.ai's Robotaxi business in Guangzhou achieved positive per-vehicle profitability. Entering the fourth quarter, according to the company's disclosure, in the recent two-week average daily operations as of November 23, Pony.ai's seventh-generation Robotaxi in Guangzhou achieved an average of 23 daily orders per vehicle, reaching positive per-vehicle profitability. The reason why "positive per-vehicle profitability" is critical is that it fundamentally answers the question of L4 autonomous driving commercialization: Can unit services generate positive cash flow? Previous calculations by Guosheng Securities show that, under neutral scenarios, Pony.ai's Robotaxi can achieve annual revenue of RMB 140,000 per vehicle, with a profit margin of about 34.9%. Achieving positive per-vehicle profitability at the city level means that L4 autonomous driving has, for the first time, validated a unit economic model in real commercial scenarios. China Merchants Securities also pointed out that the company's Robotaxi revenue is growing rapidly, with the business achieving revenue of $6.7 million in the third quarter of 2025, a year-on-year surge of 89.5%, of which passenger fare revenue increased by more than 200% year-on-year. In 2024, Pony.ai's revenue structure was 10% from Robotaxi, 54% from Robotruck, and 37% from licensing and applications. Entering 2025, this structure began to show changes in growth rates. On one hand, the Robotaxi business is accelerating its self-sustaining capability, while licensing and application revenue grew 354.6% year-on-year, continuing its high growth; on the other hand, Robotruck maintained steady growth. Pony.ai's positioning as a multi-business-form autonomous driving technology platform company has become more stable. Cost reduction through technology is the key to its increasingly viable commercialization. The BOM cost of Pony.ai's seventh-generation Robotaxi autonomous driving kit has decreased by more than 70% compared to the previous generation, with the cost of onboard computing units dropping by 80% and LiDAR costs reduced by 68%; it also plans to further reduce costs by 20% on models to be mass-produced in 2026. "The reduction in costs makes it possible to deploy larger fleets," pointed out an analyst at China Merchants Securities. By the end of 2025, Pony.ai's Robotaxi operating fleet size had exceeded 1,159 vehicles, surpassing the annual "thousand-vehicle" strategic target. This round of cost reduction and performance improvement is not accidental but stems from changes in technical pathways. Pony.ai has shifted from "imitation learning" to a training paradigm of "reinforcement learning + world models." By constructing highly realistic virtual environments, the autonomous driving system engages in massive self-play in ultra-realistic environments, forming interpretable and evolvable driving logic, thereby overcoming long-tail scenarios and achieving an effective balance between technical capability and engineering mass production. **First, Validating Robotaxi in High-End Business Travel** Just before the release of the positive update, Pony.ai also announced a major collaboration with high-end mobility service provider Aitebo to extend autonomous driving services to high-end business travel scenarios. This is a key implementation of its "light-asset + AI empowerment" strategy and a clever business tactic: establishing service benchmarks and profit models in the high-end market to accumulate momentum for subsequent penetration into the mass market. The first batch of autonomous fleets jointly formed by the two parties has obtained multi-stage license permits for L4 fully driverless testing and demonstration applications and will be connected to Pony.ai's self-operated ride-hailing platform and third-party partner platforms, covering user scenarios such as commuting, leisure, and airport transfers. High-end business travel scenarios have higher per-customer prices and stronger service standardization, providing a high-value testing ground for autonomous driving commercialization. At the same time, Pony.ai's self-operated vehicles will also be fully integrated into Aitebo's StarRide platform, building a closed loop of "order - capacity - service." Peng Jun, founder and CEO of Pony.ai, stated that the company is not only about providing technology but is also committed to exploring efficient paths for Robotaxi commercialization with partners who possess excellent service genes. Currently, its Robotaxi services cover transportation hubs such as Guangzhou Baiyun Airport, Guangzhou South Railway Station, and Shenzhen Bao'an Airport. From the perspective of the industry chain, Pony.ai is evolving into a midstream system integrator and solution provider. Its seventh-generation autonomous driving system adopts 100% automotive-grade components and platform-adaptive design, collaborating with Toyota, BAIC, and GAC Aion to launch multiple pre-installed mass-produced models. It is reported that Pony.ai's fourth-generation autonomous truck kit, to be released this year, will continue to use 100% automotive-grade components, reusing the vast majority of components from the new-generation Robotaxi autonomous driving kit, with the BOM cost per vehicle's autonomous driving kit reduced by about 70% compared to the previous generation. This collaborative model of "tech companies defining products + traditional automakers responsible for manufacturing" is one of the mainstream paths for autonomous driving commercialization in China. After achieving the thousand-vehicle target, its Robotaxi fleet size will reach 3,000 by 2026. From an industry perspective, the differentiation in Robo-X will become increasingly prominent in 2026. On one side are players still stuck in demonstration operations without a business model; on the other are companies already discussing per-vehicle profits, fleet size, and capital structure. Pony.ai's positive update does not mean it is about to turn profitable. In fact, the scalable replication of per-vehicle profitability still requires continuous validation, and as Guosheng Securities estimated, the company will still need several years to achieve full profitability, with significant potential for further cost reductions. But it clearly indicates that autonomous driving is beginning to be included in the discussion framework of "sustainable operations." Before the Robo-X track truly enters the scaling phase, the market may need a few successful model samples. In this sense, the value of this positive update may far exceed its financial numbers. ### Related Stocks - [Pony AI Inc. (PONY.US)](https://longbridge.com/en/quote/PONY.US.md) - [PONY-W (02026.HK)](https://longbridge.com/en/quote/02026.HK.md) - [Toyota Motor Corp. Sponsored ADR (N6FD.SG)](https://longbridge.com/en/quote/N6FD.SG.md) - [Toyota Motor Corporation (TM.US)](https://longbridge.com/en/quote/TM.US.md) - [Toyota Motor Corporation (TOYOF.US)](https://longbridge.com/en/quote/TOYOF.US.md)