Moonshot AI closed a $3.5 billion financing round that values the Beijing-based startup at $35 billion, well above the $1 billion to $2 billion range it had originally set as its target. The round lands a day after AI Insiders reported that Moonshot was pushing suppliers for more Nvidia chips to train its next model, and the two moves read as one strategy rather than two. Money and compute are being lined up together.
The National Artificial Intelligence Industry Investment Fund, the state-backed vehicle that has also invested in DeepSeek, was among the lead investors in the round. Its presence ties Moonshot’s raise to Beijing’s broader effort to fund domestic AI labs capable of competing with U.S. rivals despite restrictions on advanced chip exports. A state fund choosing to back a private lab at this scale is itself a data point about how much of China’s AI buildout now runs through public capital.
Investor demand for the round outpaced what Moonshot had planned to raise, a jump the company links to momentum from Kimi K3, its most recent large language model. Landing more than triple the top end of an original target range is a strong signal of investor appetite, though Moonshot has not disclosed how oversubscribed the round was in specific terms.
Moonshot is not stopping at $35 billion. The company is now approaching investors for a further round at a $50 billion pre-money valuation, which it is positioning as a final private raise before an initial public offering in Hong Kong that could arrive before the end of the year. That sequencing, one more round then a listing, suggests Moonshot wants public-market capital available on a faster timeline than most frontier labs have pursued so far.
Taken together, the oversubscribed raise, the state fund’s involvement and the pending Hong Kong listing describe a company treating capital and compute as one problem rather than two. Raising far beyond what was originally sought, accepting state money and preparing a public listing all serve the same end as chasing more Nvidia chips: securing the resources to keep training frontier models under export controls that constrain how much hardware Chinese labs can buy on the open market. Capital that arrives faster than planned buys options, including the option to pay premiums for constrained chip supply through indirect channels.
For operators tracking Chinese AI capacity, the pace at which Moonshot closes its next round and completes a Hong Kong listing will say more about its real compute trajectory than any benchmark release between now and year end.
Bloomberg reported on July 29, 2026, that Moonshot AI closed a $3.5 billion round at a $35 billion valuation with backing from China’s National Artificial Intelligence Industry Investment Fund.
