China AI start-up Moonshot reportedly eyes US$18 billion valuation in fresh fundraising push
Rapid re-rating as investors pile in
Moonshot AI (Chinese name not publicly disclosed) is reportedly in talks to raise as much as US$1 billion in an expanded funding round that would push its valuation to about US$18 billion — more than quadruple the figure from three months ago. It has been reported that the company behind the Kimi chatbot kicked off these discussions after securing more than US$700 million earlier this year at a US$10 billion valuation, and that a prior tranche late last year valued it at roughly US$4.3 billion.
Backers and the wider chase for AI dominance
Investors that have backed Moonshot previously include Alibaba Group Holding (阿里巴巴集团), Tencent Holdings (腾讯控股) and 5Y Capital (5Y资本), and those names reportedly increased their commitments at the US$10 billion level, Bloomberg and SCMP have reported. A Moonshot spokesperson did not respond to requests for comment. The speed of the re-rating underscores growing domestic appetite for AI winners as Chinese start-ups race to rival Silicon Valley names such as OpenAI and Anthropic.
Geopolitics, chips and open-source disruption
This fundraising frenzy comes against a backdrop of tighter US export controls on advanced chips and heightened geopolitical scrutiny of Chinese tech. Yet capital is flowing. Why? Partly because breakouts from the open-source ecosystem — notably the OpenClaw agent — have prompted top Chinese cloud providers and AI upstarts to launch local equivalents, accelerating adoption and investor confidence.
Stakes and questions ahead
Rivals in Hong Kong and mainland China have seen eye-popping paper valuations — some reportedly trading between US$30 billion and US$40 billion — which raises fresh questions about how quickly cash can be converted into durable products, global market share and safe, trustworthy deployments. Can Moonshot turn valuation momentum into technological lead and commercial traction? Investors seem to be betting yes, but execution will be the real test.
