The valuation landscape for Chinese artificial intelligence startups has detached from traditional metrics, entering a realm of speculative intensity that mirrors the breakneck speed of the large language models themselves. Earlier this year, Zhipu AI and MiniMax successfully debuted on the Hong Kong Stock Exchange, with Zhipu’s market capitalization surging to six times its initial value. This rally occurred despite the company generating less than one percent of the revenue of industry veterans like Baidu, signaling a market that prioritizes future potential over current solvency.
At the center of this current frenzy is Kimi, the flagship product of Moonshot AI. Reports indicate the startup is finalizing a $2 billion funding round that values the company at over $20 billion—a fourfold increase in just six months. With cumulative funding now exceeding 37.6 billion RMB, Kimi has become the most heavily capitalized private player in the domestic field, emerging as a 'last boarding pass' for investors desperate to capture the next tectonic shift in Chinese tech.
However, the dominance of established players is being challenged by the meteoric rise of DeepSeek. Once valued at a modest $10 billion, DeepSeek’s valuation has doubled repeatedly following interest from tech giants Alibaba and Tencent. Now, rumors of involvement from China’s National Integrated Circuit Industry Investment Fund (the 'Big Fund') have pushed its projected valuation toward the $50 billion mark. This state-linked interest suggests that AI is no longer just a commercial race, but a core component of national strategic infrastructure.
Kimi’s path to this peak was not linear. After a period of aggressive consumer-facing marketing in late 2024, the company briefly faded following the release of DeepSeek-R1, which redefined the market's expectations for reasoning capabilities. In response, Kimi pivoted away from expensive traffic acquisition toward algorithmic depth, focusing on coding and 'Agent' capabilities. The subsequent release of the K2 series models has successfully recaptured market confidence and driven a surge in annual recurring revenue.
Despite the influx of capital, the industry faces an existential challenge in the form of spiraling compute costs. Daily token usage in China has increased a thousand-fold over the last two years, leading to a supply-demand imbalance. Major cloud providers, including Alibaba and Tencent, have recently hiked prices for API calls, some by over 460%. This infrastructure tax means that even as startups secure multi-billion dollar rounds, a significant portion of that capital is immediately cycled back to the hyperscalers for compute power.
This capital-intensive cycle is forcing a consolidation of strategies. Leading startups are no longer attempting to build all-encompassing digital assistants to rival ByteDance’s Doubao. Instead, they are carving out niches in specialized reasoning and open-source contributions. While Kimi remains slightly behind Zhipu and MiniMax in its timeline for a public listing, the sheer volume of its latest private funding rounds suggests that for the most elite tier of Chinese AI, the public markets may be the only venue large enough to sustain their eventual appetites.