Moonshot AI, the Chinese artificial intelligence startup behind the Kimi chatbot, has officially launched its Hong Kong stock market listing process after pivoting away from expensive advertising campaigns to focus on building world-class AI models. The company submitted its A1 form, the official application for Hong Kong listing, in confidential form this week, though Moonshot AI declined to confirm the filing, stating it had "no relevant information or response to release".

What Happened to Moonshot AI's Original Strategy?

Just over a year ago, Moonshot AI seemed unstoppable. The company had launched Kimi Chat in October 2023 with a 200,000-character context window, making it one of the largest in the world at the time. By 2024, Kimi expanded that context to 2 million characters and became a viral hit, ranking in the top three AI applications in China and becoming the only "Six AI Dragons" startup to run a profitable consumer-facing AI product.

To fuel this growth, Moonshot AI spent heavily on user acquisition. In the fourth quarter of 2024 alone, the company invested 530 million yuan in advertising, with October accounting for approximately 222 million yuan of that total. The strategy worked in terms of traffic, but it masked a deeper problem: the business model was not sustainable.

Everything changed in January 2025 when DeepSeek released its R1 model. The startup achieved 30 million daily active users in less than two weeks without significant advertising spending, relying instead on word-of-mouth and superior model performance. Within two months, DeepSeek's monthly active users reached 33.7 million, nearly 1.7 times higher than Kimi's 19.43 million users at the time. The competitive landscape shifted overnight.

How Did Yang Zhilin Turn the Company Around?

Facing pressure from DeepSeek and larger tech companies, Moonshot AI's founder Yang Zhilin made two critical strategic decisions. First, he completely abandoned the consumer-focused advertising strategy that had defined the company's early years. Second, he committed the company to foundational model research and development, positioning Moonshot AI to compete on technical capability rather than marketing spend.

"It's a bit like driving on the road, there are continuous snow mountains ahead, but you don't know what's inside, you are moving forward step by step," said Yang Zhilin, describing the technical journey of building large models.

Yang Zhilin, Founder of Moonshot AI

The results have been striking. After launching K3, Moonshot AI's Series F financing round was oversubscribed by three times, with the quota closing early. The company's post-money valuation reached 35 billion US dollars. The Pre-IPO round, originally scheduled for August, was accelerated, with a pre-money valuation reaching 50 billion US dollars.

K3 has become a phenomenal achievement in the global AI landscape. The model now ranks third worldwide on the Artificial Analysis comprehensive evaluation list and is recognized as having technical strength comparable to leading proprietary models, placing Moonshot AI among the world's top-tier large model developers.

Steps to Understanding Moonshot AI's Strategic Shift

• Consumer-to-Enterprise Pivot: Moonshot AI suspended other consumer applications such as Ohai and Noisee, reducing advertising spending by more than 70% compared with the fourth quarter of 2024 and terminating partnerships with Android channel platforms and third-party advertising networks.
• Focus on Model Research: The company shifted resources entirely toward foundational model development, accepting the sunk costs of its trillion-parameter mixture-of-experts model infrastructure to compete on technical performance rather than user acquisition.
• Global Competitive Positioning: By achieving breakthrough performance with K3, Moonshot AI positioned itself alongside leading AI research organizations, attracting institutional capital that values implementable commercial performance over consumer growth narratives.

The company's transformation reflects a broader shift in how the AI industry evaluates startups. After DeepSeek's success, capital markets stopped paying for pure foundational model stories and began valuing demonstrable commercial performance instead. Moonshot AI voluntarily abandoned its consumer growth narrative, but in doing so, it positioned itself to compete on the metric that now matters most: model quality.

The IPO filing represents a milestone not just for Moonshot AI, but for the Chinese AI ecosystem. The company's journey from consumer-focused chatbot startup to world-class model developer illustrates how quickly competitive dynamics can shift in artificial intelligence, and how companies that can adapt their strategy quickly may ultimately succeed where those committed to a single path may falter.