Moonshot AI (月之暗面) raises $483 million (RMB 3.5 billion) in an oversubscribed D+ round, pushing its...
The AMW Read
Moonshot's fundraise is a significant capital event for a Chinese foundation model company, but it confirms existing patterns of valuation compression and the need for commercial narrative differentiation.
Moonshot AI (月之暗面) raises $483 million (RMB 3.5 billion) in an oversubscribed D+ round, pushing its valuation to $35 billion, and is now seeking a $50 billion pre-money valuation for a Pre-IPO round. The company's valuation has increased more than 10x since the start of 2025, when it was valued at $4.3 billion after its Series C.
This fundraise lands at a precarious moment for Chinese foundation model companies in public markets. Zhipu AI (智谱), trading at HK$1,033, has fallen 65.3% from its all-time high, while MiniMax is down 84% from its peak. The article argues that pure model capability is no longer a sufficient valuation premium — the market is now demanding proof of commercial viability and differentiated business models. Moonshot's current ~$300M ARR implies a 166x ARR multiple at its $50B target, far above the 20x-62x multiples of Anthropic, Zhipu, and MiniMax, suggesting the company needs to articulate a unique narrative beyond being "China's ChatGPT."
The capital-compression arc is becoming visible: as model capabilities converge across labs, frontier-model companies must now justify their valuations through proprietary data moats, enterprise distribution, or unique product experiences. OpenAI just raised its valuation floor to $1 trillion, Anthropic is accelerating toward an IPO, and DeepSeek has already raised at a $71 billion valuation. Moonshot's challenge is to prove it is not a clone but a distinct AI company with a defensible angle — potentially in enterprise services, given its recent open-sourcing of Kimi K3 and pivot toward B2B revenue.
