Since March, the Hong Kong stock AI sector's most explosive gainer hasn't been Zhipu or MiniMax, but rather XunCe, a company most people hadn't heard much about before. On March 5th it was trading at HK$70.65, and by March 10th it surged to HK$153.9, hitting an intraday high of HK$170.5, nearly doubling in just a few days.

On March 6th, the company released its 2025 earnings guidance, projecting full-year revenue of approximately RMB 1.283 billion, a massive 102.95% year-over-year increase.

But I think the market's most common mistake right now is rushing to call it "China's Palantir" whenever something rallies.

XunCe certainly isn't a pure concept stock—it genuinely operates in real-time data infrastructure and analytics. By prospectus standards, it ranked fourth in the Chinese market in 2023 and first in the asset management subsegment. The issue is that building a data platform and becoming a Palantir are completely different things.

What truly makes Palantir exceptional isn't just "being able to process data," but rather its ability to deeply embed itself within client organizations and ultimately take responsibility for business outcomes.

XunCe is more like a data-heavy delivery company using an AI narrative to inflate its valuation. Its story is compelling and the logic flows, but fundamentally it hasn't yet proven it has transitioned from "project-based revenue" to "platform compounding."

The crucial difference still comes down to commercialization depth. Palantir captures the most complex, most lucrative, most deeply embedded large deals; while XunCe's growth is explosive, its average deal size, customer stickiness, and platform reusability are nowhere near that level yet.