XPeng's Dual Bet: Robotaxi Tests Abroad by 2027 and a Paris Order Book Opening

XPeng targets 2027 robotaxi trials abroad and a UNECE data role, after 41,256 September deliveries and a 3.8% share gain to EUR 8.84.

XPeng is laying the groundwork to take its autonomous driving technology well beyond Chinese borders. Brian Gu, the company's vice chairman and president, confirmed on Friday at a conference in Singapore that the electric vehicle maker intends to begin initial robotaxi trials in selected cities outside China during 2027. Speaking to the news channel CNA, Gu stopped short of identifying which markets are first in line.

The fleet's hardware backbone is the GX sport utility vehicle, fitted with four in-house Turing AI chips and the VLA 2.0 software stack. According to the company, the system runs without LiDAR sensors and without high-resolution map data. Guangzhou already serves as a testing ground: since August, XPeng has held a permit for driverless trials on designated routes there, with regular passenger service in the city targeted for 2027. External partners are to handle day-to-day fleet operations, while XPeng supplies the hardware and software.

A Regulatory Foothold to Match the Commercial Ambition

Alongside the operational roadmap, XPeng is deepening its regulatory footprint overseas. On Friday it took on the coordinating role of the OICA secretariat within the newly formed UNECE task force TF-ESRI. The body, jointly chaired by France and China, exists to standardize the exchange of environmental and safety data for automated driving systems under the framework of the United Nations Economic Commission for Europe.

That regulatory work sits alongside a broader technology push. Reports indicate XPeng wants to offer its technology package to foreign automakers beyond Volkswagen, with licensing and customer-specific customization seen as potential additional revenue streams. The company has also taken a role in a UNECE initiative on international cooperation in automated driving. Such collaboration, however, is not a licensing contract, and investors should treat cooperation and commercial exploitation as separate stages.

Deliveries Provide the Operational Yardstick

The concrete measure of progress remains the vehicle business. XPeng reported 41,256 deliveries for September 2026 on October 1, a 5% increase over the prior month. For the third quarter of 2026, the company handed over 118,390 vehicles — 15% more than in the preceding quarter. Both the monthly and quarterly figures point in the same direction: volumes expanded against their respective comparison periods.

Those numbers describe completed periods, not future growth. New deliveries must follow for the next leg, and an investor who simply extrapolates the recent increase turns a positive reading into an expectation. The distinction matters: deliveries already made are a different question from revenue XPeng hopes to earn through additional licensing partners.

Robotaxis, Robots, and a Diversifying Technology Base

The mobility service is one strand of a wider diversification into robotics. Manufacturing is targeted at 1,000 units per month from the end of the current year, with more than 85% of suppliers shared with the passenger car business. That unit closed a $900 million financing round in August, drawing participation from Alibaba and Tencent, among others. External sales of the robots are planned from the second half of 2027.

A Breather for the Share Price

On the stock market, the latest developments offered a pause after a weak year. XPeng shares closed Friday at EUR 8.84, up 3.8%, edging away from a 52-week low of EUR 8.16 touched earlier in the month. The gain came amid a rally in Chinese electric vehicle names, with buying reportedly concentrated on manufacturers with sales exposure in China rather than the sector at large.

That raises a question the price move alone cannot answer: whether XPeng can back the tailwind with its own business results. Market sentiment toward Chinese autos can lift several manufacturers at once, whereas new technology revenue would be company-specific. Which of the two impulses ultimately gets support from operating results is what matters for the investment case.

The Next Scheduled Test

The next date on the calendar is October 12, when XPeng plans to unveil the G9L globally at the Paris Motor Show, open European orders, and announce European pricing. The transition from announced offering to possible demand is the part that counts. Pricing and the opening of the order books would make the European sales push more tangible, though they would still not prove later deliveries.

The favorable scenario combines two mutually reinforcing developments: XPeng sustains its delivery growth while adding further foreign manufacturers to its technology roster. The investment thesis would then no longer rest solely on rising vehicle volumes. The decisive element in the reported plan is the expansion beyond Volkswagen — not the offer to additional automakers itself, but its conversion into concrete business relationships. Only that could generate the intended extra revenue contribution.

The sober counter-image is a recovery carried mainly by demand for Chinese auto stocks. Should that market impulse fade, company-specific progress would have to carry more weight. The reported licensing plans offer a perspective, but not yet a basis for treating additional revenue as already earned. The risk lies in valuing both impulses too quickly: demand-driven buying can move the stock without producing an additional order for XPeng, while a promising technology offensive may need time for commercial execution. The weaker scenario would be a share price that operating business fails to follow.