According to Reuters, Moonshot AI has confidentially filed for a Hong Kong IPO that could raise about $3 billion, marking a significant escalation from its previous funding rounds. The Beijing-based startup is currently valued at $50 billion in a funding round still in progress and has raised more than $5.5 billion since 2023. This IPO filing comes as the company continues to target a manifold jump in annualized revenue to $2 billion by the end of 2026, using the breakout success of its Kimi K3 model to compete with rivals.
As reported by Reuters, Moonshot is in early-stage talks with Microsoft Corp. (NASDAQ:MSFT), Amazon.com Inc. (NASDAQ:AMZN) and Alphabet Inc. (NASDAQ:GOOGL) over deals that could bring its models deeper into U.S. cloud platforms. The company is seeking as much as 30% of K3-related revenue through revenue-sharing agreements for Kimi K3. Any agreement could mark the first major revenue-sharing pact between a Chinese AI developer and a major U.S. cloud provider. This strategic move addresses the challenge of running the massive Kimi K3 model, which has 2.8 trillion parameters and makes it expensive to run, helping explain why demand after its July launch strained Moonshot's computing capacity.
According to BeInCrypto, Moonshot released the world's largest free artificial intelligence model, Kimi K3, on July 16, allowing anyone to download and run it. However, the launch created unprecedented demand that quickly overwhelmed the company's infrastructure. Four days into the launch, Moonshot stopped selling new subscriptions as paying members kept their seats while everyone else waited. The company stated that "over the past 48 hours, demand has pushed close to the limits of our current capacity." On September 11, Moonshot pushed a smaller model, K2.8 Preview, across its coding platform, with some K3 requests now being answered by the smaller model instead.
According to BeInCrypto, Anthropic accused Moonshot of covertly passing Kimi users' questions to Claude models and passing off the responses as Moonshot's own work. In a single 10-day window, Anthropic counted nearly 300,000 handoffs routed through 5,380 fake accounts. The allegations include claims that one of those users was probably military, who uploaded footage from hundreds of security cameras in Chengdu and asked whether a man was behaving strangely. On Saturday, Moonshot AI responded by stating that "the information circulating online regarding the founder and employees is completely fabricated and malicious slander." The company immediately reported the matter to the police and will pursue legal action against those responsible for spreading the rumors.
According to reports from NDTV, Moonshot is targeting a manifold jump in annualized revenue to $2 billion by the end of 2026, using the breakout success of its Kimi K3 model to compete with rivals. The Beijing-based startup recently told investors that its annual recurring revenue — a gauge of forward-looking sales — topped $1 billion in August, up from $300 million in June. This dramatic leap was fueled by the release in July of the massive Kimi K3 artificial intelligence model, which leapt to the top of performance benchmarks while costing drastically less than top US contenders.
According to NDTV, Moonshot is in the process of raising funds at a $50 billion valuation, roughly in line with Z.ai's current market value, ahead of an initial public offering in Hong Kong as soon as this year. However, Chinese AI startups face constant scrutiny over their technical capabilities, with Moonshot and domestic rivals facing allegations of illicit distillation from top US labs. Anthropic this week accused Moonshot of covertly routing user requests to its Claude models and passing off the responses as Moonshot's own, an allegation Moonshot hasn't yet publicly addressed. The broader context includes ongoing US-China tensions over AI technology, with export controls creating additional challenges for cross-border partnerships.
