In just five months, a company has secured financing totaling 267 billion yuan.
Recently, whether scrolling through tech media or social feeds, one recurring piece of news is unavoidable: Kimi has raised another funding round.
In January 2026, it closed a round. In February, another. In May, yet another—this time a $2 billion round, valuing Beijing Moonshot AI Co., Ltd. (Kimi's parent company, hereinafter "Kimi") at $20 billion post-money, approximately 1,367 billion yuan. In less than half a year, the company has amassed over $3.9 billion (about 267 billion yuan) in cumulative financing.
To put that in perspective, it averages nearly 200 million yuan flowing in daily. This speed and scale are unmatched in the domestic large language model (LLM) arena.
But more intriguing than the numbers is who is providing the capital. Kimi's shareholder roster includes industry giants like Sequoia, IDG, and ZhenFund, tech titans such as Alibaba and Tencent, and, notably, state-backed entities: Guozhitou, the Beijing Artificial Intelligence Fund, and China Mobile. An AI LLM startup merely three years old has clearly captured the favor of diverse capital sources.
Behind the company is founder Yang Zhilin—a Tsinghua graduate with a Ph.D. from Carnegie Mellon, fitting the "technical prodigy" archetype at just 34 years old. He once stated publicly, "We're in no rush to go public; let's focus on perfecting the technology first." Yet, within China's LLM landscape, Kimi's fundraising pace has visibly accelerated. What is compelling this sprint?
01# "Tsinghua Prodigy" Becomes Capital Darling Kimi's Valuation Soars in Six Months
Within the LLM circle, Yang Zhilin is considered a "prodigy among prodigies." Born in 1992, he graduated top of his class from Tsinghua's Computer Science department under the mentorship of Tang Jie (a foundational figure in China's LLM sphere). His doctoral advisor at Carnegie Mellon was Apple's AI head. Before returning to China to start his venture, he was widely recognized as "one of the few individuals in China who truly understands language models."
In 2023, upon witnessing ChatGPT, he judged the "window of opportunity might only last a month," promptly founding Moonshot AI and launching its core product—the AI assistant Kimi. Simply put, Kimi can read and comprehend lengthy documents like theses, contracts, and research papers spanning hundreds of thousands of words at once, providing accurate Q&A and summaries. It also efficiently writes, modifies, and debugs code.
Leveraging exceptional long-context processing capabilities and robust code-generation functions, Kimi quickly distinguished itself in the AI market.
In March 2026, tech blogger Avi Chawla posted a detailed thread on X dissecting a recent technical report released by the Moonshot AI Kimi team. Shortly after the post went live, Elon Musk himself replied beneath it, stating: "Impressive work from Kimi."
Subsequently, Moonshot AI embarked on a continuous fundraising spree, with its valuation climbing steadily. In April 2023, merely two months after incorporation, the company completed an angel round of nearly 2 billion yuan, attracting top-tier investors like Sequoia China and ZhenFund, achieving a $300 million valuation. By July of the same year, it secured a Series A round with participation from Meituan Longzhu and Lanchi Ventures.
In February 2024, it raised over $1 billion, with Alibaba contributing approximately $800 million, propelling the company's valuation to $2.5 billion. Tencent joined in August 2024, raising the valuation to $3.3 billion.
The latter half of 2025 saw Moonshot AI experience a fundraising lull of over a year. This ended in December when the company closed a $500 million Series C round led by IDG, with existing shareholders like Alibaba and Tencent participating in oversubscription, resulting in a post-money valuation of $4.3 billion.
According to reports from the new media outlet "AI Business Technology," by late 2025, the company was in talks with IDG Capital, Tencent, and others for a new funding round potentially valuing it between $3.8 and $4 billion, with plans to initiate an IPO in the second half of 2026.
However, on New Year's Eve, December 31, 2025, Yang Zhilin sent an internal letter to all employees, writing: "Compared to the secondary market, we believe we can still raise significantly more capital from the primary market. In fact, the amounts raised in our Series B/C rounds already exceed the fundraising of most IPOs and follow-on offerings by listed companies."
He further noted, "Therefore, we are in no hurry to go public in the short term. Of course, we plan to use listing as a means to accelerate AGI in the future, timing it appropriately, with the initiative firmly in our hands."
Scarcely had these words settled. From January to February 2026, the company consecutively closed three funding rounds of $500 million, $700 million, and $700 million respectively. Its valuation surged from $10 billion to $18 billion, with existing shareholders like Alibaba, Tencent, and 5Y Capital continuing to invest.
In May, a $2 billion infusion arrived—the largest single funding round since Moonshot AI's inception—pushing the post-money valuation past $20 billion, approximately 140 billion yuan. This round was led by Meituan Longzhu, with participation from Tsinghua Capital, China Mobile, CPE, and others. Within six months, the valuation skyrocketed from $4.3 billion to $20 billion.
During this period, market rumors about a potential IPO for the company proliferated. On May 8, 2026, according to "First Finance" reports, Moonshot AI had initiated discussions for a Hong Kong IPO, evaluating timing windows and structural plans while assembling an上市 team.
It is noteworthy that the early shareholder base primarily consisted of internet giants like Alibaba, Tencent, and Xiaohongshu, alongside leading VCs such as Sequoia China, IDG Capital, and ZhenFund. Following the latest round, state-backed platforms and central enterprise giants like Guozhitou, the Beijing Artificial Intelligence Fund, and China Mobile formally entered the core shareholder ranks.
From the public stance on December 31, 2025, of "not aiming for listing as a goal," to May 2026, when market reports emerged of密集 fundraising,启动 restructuring for a红筹 listing, and a sprint toward a Hong Kong IPO, Yang Zhilin's shift in attitude took less than five months.
02# Peers Take Flight, AI Window Closes Fast Kimi Finds Itself in IPO Golden Period
The answer to this转变 may lie within that early 2026 window.
In January 2026, Zhipu AI and MINIMAX-WP successively listed on the Hong Kong Stock Exchange. Zhipu AI (2513.HK) had an IPO price of HK$116 per share, later soaring above HK$1,000, with its market cap一度突破 HK$400 billion. MINIMAX-WP (0100.HK) debuted at HK$165 per share, closing with a market cap of HK$106.7 billion.
As of May 22, 2026, Zhipu AI's stock price closed at HK$1,282 per share, with a market cap of HK$571.57 billion; MINIMAX-WP's stock price closed at HK$768.5 per share, with a market cap of HK$241.03 billion.
Early-stage funds saw paper returns multiply by hundreds, instantly transforming the AI赛道 into a "miss-it-once, regret-it-for-a-lifetime" opportunity.
Compared to its peers, Kimi's own commercialization has proven viable. According to "The Paper" reports, by late February 2026, less than a month after the release of the Kimi K2 Thinking model, its cumulative revenue over nearly 20 days had already surpassed the total for the entire year of 2025. In March, its Annual Recurring Revenue (ARR)突破 $100 million. By April, that figure doubled to 200 million yuan.
Senior industry commentator Zhang Shule believes that Kimi's powerful long-text reading comprehension capability solves the LLM problem of "losing coherence in longer conversations." This represents a compelling narrative highly favored by capital markets and a critical breakthrough urgently needed in AI, one that has already been validated by users.
High赛道 sentiment, soaring peer stock prices, and proven自身 commercialization—triple叠加—make the present moment the peak of Kimi's valuation.
Simultaneously, the burn rate for LLMs far exceeds expectations. According to "Sohu Tech" reports, the training cost for the K2.5 model is approximately $4.6 million (about 32 million yuan). The company's budgeted computing power expenditure for this year exceeds $1.5 billion.
Zhang Shule points out that AI competition shifted from an LLM arms race to an agent competition in 2025. While many peers focused on单项赛事, Kimi quietly completed a dual-track布局 of "LLM + Agent智能体集群," requiring further capital to "solidify its foundation" and achieve "ascent." Whether it's accelerated fundraising, valuation explosion, or advancing toward an IPO at this moment, these moves顺势而为, driven by潮流 rather than资本倒逼.
Shen Meng, Director of Chanson & Co., offers another perspective from the external environment. He notes that expectations are that both OpenAI and Anthropic will complete IPOs within the year, so major Chinese LLM enterprises will attract more investor attention and command better valuations during the same period.
Looking back at Yang Zhilin's internal信 phrase "in no hurry to go public in the short term," it was not a rejection of listing but rather waiting for the optimal时机.
Facts have borne this out. Less than a month after the internal信 was sent, three funding rounds密集落地, with valuation飙 from $4.3 billion to $18 billion. According to "First Finance" reports on May 8, Moonshot AI had initiated discussions for a Hong Kong IPO, evaluating timing windows and structural plans while recruiting an上市 team.
This marks the company's IPO process entering a substantive推进 stage, positioning it as another top contender冲击 the Hong Kong stock market following Zhipu AI and MINIMAX-WP within China's AI LLM赛道.
Shen Meng补充道 that investors关注 the valuation development cycle of enterprises, hoping to exit at market highs. Currently, AI sits in a middle ground between rational narrative and overheated泡沫, which恰好 is the golden区间 most suitable for冲击 an IPO.
Yang Zhilin's most recent public appearance was on April 10, 2026. He participated as an entrepreneur representative in an economic形势专家和企业家座谈会. Among the eight invited guests, he was the sole representative from an independent LLM company and the youngest attendee.
This race is far from decided. Shen Meng believes that, regardless, the LLM market will ultimately trend toward consolidation where the big get bigger. Capital will subsequently seek new opportunities along the entire AI industry chain. Future industry整合 will compete not only on technology and products but also on talent integration.
Zhang Shule直言 that the LLM赛道 has not yet passed the "college entrance exam"关卡; only upon entering the "graduate student"阶段 can true胜负 be determined. Opportunities for颠覆式创新依然存在.
Five months, 267 billion yuan, a $20 billion valuation. The technical prodigy who three years ago果断创业 upon judging a "window of opportunity might only last a month" now stands before a much larger window. This time, what he must do is not merely seize it.