A commuter in Suzhou can hail a self-driving car, but the business that pays the engineers is more likely to be the driver-assistance software in a friend's family sedan. That distinction is the key to understanding Momenta, the Chinese autonomous-driving developer that moved toward a Hong Kong listing in mid-2026.
Reuters reported in June 2026 that Momenta was planning to launch its initial public offering, potentially raising about US$900 million at a valuation near US$9 billion. The figures came from people familiar with the plans and were tentative; the company told Reuters it was not in a position to comment on market speculation. What is solid is the underlying question investors were being asked to answer: can a company sell driving software at scale before fully driverless taxis become profitable?
Key takeaways
- Reuters reported in June 2026 that Momenta sought to raise about US$900 million in a Hong Kong IPO at a valuation near US$9 billion, with a timetable that remained subject to change.
- The company was founded in 2016 by Cao Xudong, a former Microsoft Research employee, and is headquartered in Suzhou.
- Its IPO filings showed losses widening to ¥3.5 billion (≈ US$490M / HK$3.8B) in 2025 from ¥2.6 billion (≈ US$364M / HK$2.8B) two years earlier.
- Backers include Toyota, Mercedes-Benz, SAIC, General Motors, Bosch, Temasek, Yunfeng Capital and Tencent.
- Momenta's revenue comes mainly from supplying driver-assistance systems (辅助驾驶) to automakers, not from robotaxi fares.
Two businesses, one balance sheet
Momenta has always had two faces. One sells production-ready driver-assistance software to carmakers — the technology that helps a car hold its lane, change lanes on a highway and navigate city streets under human supervision. The other runs paid robotaxi (无人驾驶出租车) services in selected Chinese cities and develops fully driverless systems.
The first business is commercially real. Reuters noted that Momenta's advanced driver-assistance systems are used by global brands including Toyota, Mercedes-Benz and Audi in China. Reporting on the filings put 2025 revenue at ¥2.41 billion (≈ US$337M / HK$2.63B), roughly 82 percent higher than a year earlier, driven by licensing and technical services. That is a company selling into mass-production (量产) vehicles, where a design win can cover hundreds of thousands of cars.
The second business is a cost center for now. Reuters reported that losses widened to ¥3.5 billion (≈ US$490M / HK$3.8B) in 2025 from ¥2.6 billion (≈ US$364M / HK$2.8B) in 2023. Running a robotaxi fleet is expensive: vehicles need mapping updates, remote supervision and, in early phases, safety drivers. The software business subsidizes the ambition.
That structure is why the IPO is interesting. Investors are not simply betting on driverless taxis arriving soon. They are buying a supplier with existing automaker revenue that also owns an option on a future network.
Why carmakers are on the shareholder list
The investor roster reads like a roll call of the auto industry. Reuters listed Toyota, Mercedes-Benz, SAIC and General Motors among existing backers, along with Bosch, Temasek, Jack Ma's Yunfeng Capital and Tencent. Prospective cornerstone investors mentioned in the reporting included Mercedes-Benz, Boyu Capital and Oaktree Capital Management.
There is a commercial logic to automakers owning a piece of their supplier. It gives them early access to the technology and a voice in priorities, while giving Momenta long-term customers. It also creates a circularity that analysts watch: when a customer is also a shareholder, a purchase is both revenue and an investment decision.
Momenta's international plans add another layer. The company has a research center in Stuttgart and said it would begin Level 4 autonomous-vehicle tests in Germany in 2026 with Uber. Its systems are already described as used by global brands in the Chinese market. For a supplier, that is the harder and more valuable kind of globalization: being designed into a car rather than merely tested in a city.
The listing environment
The offering came during a busy period for Hong Kong. Reuters cited LSEG data showing Hong Kong IPOs, including secondary listings of mainland-listed companies, raised US$21.6 billion in the first half of 2026, up 51 percent from a year earlier. Momenta had earlier considered a New York listing before shifting to Hong Kong, according to people familiar with the plans.
For readers, the macro detail matters less than the signal: capital markets were being asked to fund a long-duration technology build at a time when investors had become more demanding about near-term revenue. Momenta's answer was to lead with mass-production revenue and treat robotaxis as the future. Whether that framing holds depends on execution, not on the pitch.
What the investment story says about ordinary roads
The strange thing about autonomous driving is that its most visible symbol — an empty driver's seat — is not yet its main product. The technology reaching ordinary buyers first is assistance: a system that reduces the routine burden of driving but still requires an alert human.
That has consequences for how the public should discuss it. A company's valuation may be justified by a driverless future, while the product in your neighbor's car today is a supervised assistance system. Both can be true, but they should not be conflated. When a manufacturer says "autonomous," the useful question is what the driver is still expected to do.
There is also a cost curve to watch. Assistance software sold to automakers improves through data collected from large fleets. Each production vehicle contributes examples of awkward junctions, poor weather and unusual road users. That is why volume matters so much in this industry, and why a supplier with hundreds of thousands of vehicles on the road has a different position from a startup running a small pilot fleet.
Honest limitations
This is an analysis of a reported offering, built on Reuters' June 2026 reporting and on financial-press summaries of Momenta's filings. The offering size, valuation, timetable and cornerstone names were reported as tentative and may have changed; no claim is made here that the listing was completed on those terms. The revenue figure comes from reporting on the filings rather than direct review of the prospectus. Loss figures are attributed to Reuters' reading of the filings. The article makes no prediction about the company's profitability, the safety of its systems, or any regulatory decision, and it avoids commentary on government policy.
What readers can do now
- If you are considering a car with advanced driver assistance, ask whether the system requires supervision and what happens if you stop paying for a software subscription.
- When you read a robotaxi headline, check whether the story describes a paid service, a test with a safety driver, or a company valuation.
- For investors, separate a company's current revenue source from the future business used to justify its valuation; autonomous driving is a case where the gap can be large.
