China's Qujing raises $195M Series B as AI inference bets surge
What's the deal? Beijing Qujing Technology, which builds AI inference infrastructure under the brand Approaching.AI, has raised US$178.8M in a Series B round announced in October 2026. The round drew a crowded investor roster led by Zhejiang Finance Capital, Xiaoshan Capital, ICBC Investment, Zhongguancun Science CityDealroom has a profile for this one. Try Dealroom →, Xinglian Capital, and Qianjiang Century City.
Who else joined? Participants included Yunchuang Capital, Hunan Hi-Tech Zongheng, CCB Huake, Shenwan Innovation InvestmentDealroom has a profile for this one. Try Dealroom →, CITIC Juxin, SenseTime, and Wangsu Science & TechnologyDealroom has a profile for this one. Try Dealroom →, among others. Existing backer Shangshi Capital added to its stake.
Why now? The deal caps an unusually fast fundraising run: Qujing closed a Pre-A round in May 2026 and an A round in July, lifting its total for the year to US$330M. It serves large-model developers including Zhipu and Moonshot AI.
How big is this? At roughly $195 million, the round ranks near the top of Chinese enterprise software Series B deals — above the 95th percentile across 301 comparable rounds.
The context: Capital is flooding into AI inference and token-serving providers. SiliconFlow raised more than US$275M in June, and Infinigence closed US$316.3M in October, as investors bet that spreading large-model applications will drive demand for inference.
What could go wrong? Qujing has not disclosed revenue, profitability, or per-inference cost figures. The round signals sector heat and scaling ambition, but the funding alone does not prove the business model works.
The signal: The scramble to back inference infrastructure reflects a shift in China's AI market — from training frontier models to monetising the compute that runs them at scale.
Read more: sohu.com
Image credit: cbowns
