Zhipu’s stock jumped 33 percent Monday, capping a tenfold rise since its Hong Kong IPO. The Chinese AI firm is no longer the underdog but the market leader, overtaking MiniMax by nearly three times in valuation.

This rally follows the U.S. government’s clampdown on Anthropic, forcing withdrawal of its advanced Claude AI models from foreign users. Zhipu capitalized immediately, announcing the release of GLM-5.2, its newest model, as open-source software. No restrictions, no strings.

The message was clear: AI power shouldn’t be a gated club. Zhipu’s approach contrasts with the U.S. trend of tightening access. For enterprises and developers hungry for affordable, capable AI, China’s open models have become the viable alternative.

Wall Street banks noticed. JPMorgan raised Zhipu’s price target to HK$1,400 from HK$950 and dropped MiniMax’s rating. Bank of America started coverage on both with buy calls, betting on China’s cost-conscious market to steal ground from locked-down U.S. rivals.

GLM-5.2 reportedly matches Anthropic’s Claude Opus 4.7 in code generation and long-context tasks during early tests. A one-million-token context window is not just a gimmick; it’s a feature tailored for developer needs and enterprise workflows.

Behind the hype, Zhipu’s financial growth is undeniable. The company posted 2025 revenue of 7.24 billion yuan, up 132 percent year-over-year. Yet its R&D spending dwarfs revenue at 31.8 billion yuan, signaling a costly race to stay competitive. Profitability may remain out of reach until 2029.

Zhipu’s market cap peaked near 1 trillion yuan, fueled by soaring API usage—up 400 percent—and pricing hikes of up to 83 percent. But looming shareholder lockup expirations could trigger volatility. Investors brace for waves of selling in the coming months.

China’s AI scene is crowded and cutthroat. Zhipu competes fiercely with MiniMax, DeepSeek, and Qwen, especially in coding and AI agent markets. China’s fragmented landscape and price-sensitive customers keep pressure on margins.

Talent migration adds another twist. Almost 40 percent of U.S.-based AI engineers were born in China. U.S. restrictions on Anthropic risk pushing this talent back home, helping Chinese firms scoop up expertise and deepen their moat.

Ultimately, Zhipu’s rise is a symptom of broader geopolitical shifts. Every U.S. door shut on AI access drives users and engineers toward China’s open models. The battle for AI dominance is no longer just about technology—it’s a turf war of policy, talent, and market access.