By Jiemian News
On May 7, according to Huafeng Capital, Kimi, a leading domestic large model company, has completed a new round of financing of approximately $2 billion, with a post-investment valuation exceeding $20 billion. This is the largest single-round financing since Kimi’s establishment and one of the highest private financing amounts for domestic large model startups.
This round of financing was led by Meituan Dragon Ball, with participation from Waterwood Capital, China Mobile, and CPE (CITIC Industrial Fund). Huafeng Capital served as the financial advisor for part of the buyers in this transaction, providing key support in transaction structure design, investor matching, and negotiation.
In less than half a year, Kimi has completed four rounds of financing, with a cumulative amount exceeding $3.9 billion, surpassing its peers, such as Zhijue AI and MiniMax, to become the leading domestic large model startup. Compared to its valuation of approximately $4.3 billion in November 2025, its current valuation has more than quadrupled.
Kimi was founded in April 2023 by Yang Zhi-Lin, a Tsinghua University undergraduate and Carnegie Mellon University Ph.D. Yang previously worked at Google Brain and Meta AI and is a top researcher in the NLP field. Just two months after its establishment, Kimi completed an angel round of financing worth nearly $200 million, with investments from top institutions such as Sequoia China and ZhenFund. In October 2023, Kimi launched Kimi Chat, which supports 200,000 Chinese characters and has established a differentiated competitive advantage in long-text processing.
In February 2024, Kimi completed a financing round exceeding $1 billion, with a valuation of $2.5 billion. In August of the same year, it received another $300 million in financing, increasing its valuation to $3.3 billion. In December 2025, Kimi completed a $500 million Series C financing round, with a valuation of $4.3 billion and a cash reserve exceeding $10 billion.
In the first two months of this year, Kimi completed three consecutive financing rounds of $500 million, $700 million, and $700 million, with its valuation rising from $10 billion to $18 billion. Existing shareholders, including Alibaba, Tencent, and Wuyuan Capital, continued to increase their investments.
With this latest financing, Kimi’s cumulative financing amount has exceeded $3.76 billion, making it the domestic large model startup with the most cumulative financing. In comparison, MiniMax has a cumulative financing amount of approximately $1.5 billion, and Zhijue AI has approximately $1.3 billion. As of the time of publication, MiniMax’s market value is approximately $22.07 billion, and Zhijue AI’s market value is approximately $36.27 billion.
Huafeng Capital believes that Kimi’s current valuation of approximately $14 billion still has significant growth potential. This round of financing will further solidify Kimi’s leading position in model SOTA, computing power reserves, and talent incentives.
Technological iteration and commercialization explosion have jointly supported Kimi’s high valuation. On April 20, Kimi released and open-sourced the Kimi K2.6 model, which is the company’s strongest code model to date. The model’s long-range encoding ability has been significantly improved, and its autonomous execution capability has been greatly enhanced. Kimi K2.6 is now available on kimi.com, the latest version of the Kimi app, Kimi API, and Kimi Code programming assistant.
In March, news reported that Kimi K2.5 model was released a month ago, and Kimi’s ARR (annual recurring revenue) exceeded $100 million. According to Huafeng Capital, Kimi’s ARR has exceeded $200 million in April, with paid subscriptions and API revenue accelerating growth.
Along with the revenue explosion, the market has frequently rumored that Kimi is about to go public. However, Kimi’s founder, Yang Zhi-Lin, has shown a calm attitude towards the IPO.
In an internal letter published at the end of last year, Yang Zhi-Lin stated that the company’s B-round and C-round financing amounts exceeded the IPO fundraising and directed issuance of most listed companies. “Therefore, we are not in a hurry to go public in the short term. Of course, we plan to go public in the future to accelerate AGI, and we will take action at the right time, with the initiative in our hands.”
Currently, the domestic large model field has formed three major camps: internet giants such as ByteDance and Alibaba, which rely on their ecosystems and computing power to comprehensively suppress; Zhijue AI and MiniMax, which have listed on the Hong Kong stock market and gained capital market advantages; and new players like DeepSeek, which have rapidly risen with open-source and low-cost strategies.
Kimi has long-text processing as its core barrier, and Kimi leads in knowledge work scenarios. However, long-text processing has gradually become an industry standard, and the technical moat is continuously narrowing. In terms of commercialization, the company’s consumer-end user growth is under pressure, and its enterprise-end customization and ecosystem construction are not as good as those of giants and some peers. The company’s revenue structure is single and risky, relying too heavily on API and subscription models.
Furthermore, high computing power costs and intense talent competition still pose challenges to Kimi’s long-term development.
On the same day that Kimi’s new financing news was announced, it was revealed that the National Integrated Circuit Industry Investment Fund is negotiating to lead DeepSeek’s first external financing round, with a post-investment valuation expected to reach around $45 billion. However, a source said that the final valuation has not been determined, and the final investment list and transaction terms are still under negotiation. Analysts believe that DeepSeek’s willingness to open up external financing is to expand its computing power resources and retain top technical talent.
In February last year, Kimi released the reasoning model K1.5, which was overshadowed by DeepSeek-R1, which was released and open-sourced on the same day. Subsequently, Kimi decided to significantly reduce its product launch budget, including suspending multiple Android channels and third-party advertising platform cooperation, and temporarily suspending “burning money.” Before DeepSeek’s rise, Kimi was the domestic AI product with the highest topic value at the time.
As major large model companies are strengthening their technical moats, Kimi urgently needs to supplement its financial ammunition and prevent core technical personnel from being poached by competitors.
