$XPeng(XPEV.US)

Recorded XPeng's VLA2.0 on February 28th.

Discussions about VLA2.0 have been increasing over the past two days. The entire Chinese autonomous driving industry is waiting for He Xiaopeng!

Currently, VLA2.0 is in the promotion and marketing phase. Capabilities are being gradually revealed. I've collected some short videos I came across.

❶ L4 testing, driverless, campus roaming. With no one in the driver's seat and no map of the campus, it completed a full set of operations including starting from a standstill, autonomous driving, pulling over temporarily, picking up passengers, and starting again.

❷ Actively yielding to ambulances

❸ Navigating narrow construction zones at night

2026/02/14

❹ International car reviewer Hammond's first experience with Chinese autonomous driving: XPeng's second-generation VLA

❺ Personal take: XPeng's VLA2.0 is taking a disruptive innovation path. It has a high ceiling but also a high floor. As the version continues to be optimized, it will become smarter and more user-friendly. However, the key is how to market it to break through the noise. I've seen many reviews so far. What gets published is naturally the positive side, but users will scrutinize any weaknesses with a magnifying glass. The challenge is how to ensure the first batch of U-version users can use it with confidence while also attracting curious potential users to learn about and experience it firsthand. One thing XPeng is doing well is simultaneous L4 testing. If they can market it as achieving L4, that's a qualitative leap compared to L2 and L3. Users can immediately see the difference and the impressive aspects, which sparks interest, leading to conversion through potential seeding. I'm not worried about its practicality. The real challenge is how to break through in marketing!

❻ On the investment side: Still holding a short position, waiting for February sales figures to gauge the trend.