XPeng Plans 2025 Robotaxi and Humanoid Robot Launches with In-House AI Chips

• XPeng’s robotaxis use four Turing chips for advanced vision-language-action models, enabling fully autonomous driving without human input.

• Testing begins in Guangzhou in 2026, expanding to other cities with external displays for pedestrian safety.

• The second-generation Iron humanoid robot enters mass production next year, powered by three Turing chips and solid-state batteries, targeting commercial applications first.

Discover XPeng’s bold move into robotaxis and humanoid robots at its 2025 AI Day. Learn how in-house chips and partnerships drive autonomous innovation. Stay ahead in EV tech—explore now for key insights on future mobility.

What is XPeng’s Plan for Launching Robotaxis?

XPeng robotaxi launch strategy centers on deploying three models in 2026, each equipped with four in-house Turing AI chips delivering 3,000 TOPS of computing power. This initiative, announced at the company’s AI Day event in Guangzhou, represents a pivot from earlier statements downplaying near-term robotaxi viability toward building comprehensive autonomous systems. The vehicles operate on a second-generation vision-language-action (VLA) model, processing visual inputs to perform actions independently, enhancing safety and efficiency in urban environments.

How Does XPeng’s Humanoid Robot Development Integrate with Its Autonomous Tech?

XPeng’s second-generation Iron humanoid robot, set for mass production in 2026, leverages three Turing AI chips and solid-state batteries for robust performance in commercial settings. CEO He Xiaopeng emphasized that these robots will initially serve as tour guides, sales assistants, and facility navigators, starting in XPeng’s own operations, due to current high costs and China’s competitive labor market. The integration with autonomous vehicle tech stems from shared AI foundations, allowing the robot to interpret environments similarly to the robotaxis’ VLA model. According to XPeng’s disclosures, this unified approach could lead to higher production volumes for robots than cars over the next decade, underscoring the company’s expansive vision in AI-driven mobility. Expert analysis from automotive technology reports highlights how such in-house chip development reduces dependency on external suppliers, potentially accelerating deployment timelines compared to competitors relying on off-the-shelf solutions.

Frequently Asked Questions

What Partnerships Support XPeng’s Robotaxi Operations?

XPeng collaborates with Alibaba, utilizing AutoNavi for mapping and the Amap ride-hailing platform for routing and bookings. This integration embeds the robotaxis into an established ecosystem, avoiding the need for a standalone service and enabling efficient scaling. Testing in Guangzhou from 2026 will leverage this for real-world validation, with expansion planned to other Chinese cities, as per company announcements.

How Does XPeng’s Driver-Assist System Compare to Competitors?

XPeng’s upcoming driver-assist system, launching in Q1 2026, minimizes manual inputs for navigating narrow roads, ideal for dense urban areas like European cities. It outperforms Tesla’s Full Self-Driving in test scenarios, completing routes faster with less human intervention, according to XPeng executives. Volkswagen will adopt it first, with plans to extend to other manufacturers, building on XPeng’s early 2023 software releases in China that have gained widespread traction in the EV sector.

Key Takeaways

• Shift to Autonomous Focus: XPeng is evolving beyond electric cars to full-stack AI systems, with robotaxis launching in 2026 using proprietary Turing chips for 3,000 TOPS performance.
• Humanoid Robot Commercialization: The Iron robot’s mass production next year targets non-residential uses initially, powered by advanced batteries and customizable features for diverse applications.
• Strategic Partnerships and Expansions: Integration with Alibaba enhances robotaxi viability, while new driver-assist tech positions XPeng as a leader, offering it to partners like Volkswagen for broader market impact.

Conclusion

XPeng’s AI Day revelations position the company at the forefront of autonomous robotaxi and humanoid robot innovation, powered by in-house Turing chips and strategic alliances like that with Alibaba. By addressing urban mobility challenges through advanced VLA models and driver-assist systems superior in efficiency to rivals, XPeng demonstrates deep expertise in AI integration for vehicles and robotics. As testing ramps up in 2026, these developments signal a transformative era in transportation—investors and tech enthusiasts should monitor XPeng’s progress for opportunities in the evolving autonomous landscape.

XPeng’s commitment to self-reliant technology underscores its competitive edge in China’s dynamic EV market. The robotaxis, with their external safety displays and high-compute capabilities, aim to redefine ride-hailing by seamlessly blending into existing platforms. Meanwhile, the humanoid robots’ focus on commercial viability reflects pragmatic scaling, avoiding premature home entry amid cost barriers. Co-President Brian Gu noted similarities to Tesla’s pursuits but highlighted XPeng’s earlier starts in areas like flying car prototypes, though with quieter publicity until now. This balanced approach, combining innovation with measured rollout, could yield significant long-term gains.

Further details from the event reveal XPeng’s testing protocols, starting localized in Guangzhou before nationwide expansion, ensuring regulatory compliance and data-driven refinements. The driver-assist system’s emphasis on narrow-road navigation addresses real-world pain points, particularly in global markets beyond China. With Volkswagen’s adoption, XPeng not only validates its tech but also fosters industry-wide advancements. Overall, these announcements affirm XPeng’s trajectory toward diversified AI applications, potentially outpacing traditional automakers in autonomous adoption rates.