Sand.ai, which builds video-generation models, announced a $200 million Series A+ round on 5 November and said it now counts as a unicorn, Chinese outlet Jiemian reported (Jiemian, in Chinese). The company said it was its third round of 2026, bringing its fundraising over the past ten months to $300 million.
By our arithmetic, the new round is 67% of that total, which leaves $100 million for the two earlier 2026 rounds combined. Sand.ai did not break out those two rounds.
Business Pill · Power Law Explained
A short explainer on power laws, the pattern behind outsized venture outcomes. It teaches the general idea only and does not discuss this company.
The key insight: Sand.ai says its $200 million Series A+ brought its 2026 fundraising to $300 million across three rounds; by our arithmetic the new round is 67% of that total, and the valuation above $1 billion is a company-supplied figure.
Jiemian reported that the round was co-led by Cornerstone Capital (基石资本) and Shenzhen-listed Wangsu Science & Technology (300017.SZ). It listed further participants including other venture firms, investors with state backing in Shandong and Shanghai, and existing shareholders adding to their stakes.
TechNode Global reported that Wangsu’s exchange disclosures describe Sand.ai as a developer of video-generation models and applications, and that the two companies intend to work together on audio and video inference, low-latency edge inference and enterprise deployment (TechNode Global).
Sand.ai’s $200 million Series A+ against the $100 million it raised in its two earlier 2026 rounds combined. Company-supplied figures via Jiemian and TechNode, 5-6 November 2026; the $100 million figure and the 67% share are our arithmetic.
The Valuation Claim
Sand.ai said the round pushed its valuation above $1 billion. TechNode Global notes that “The valuation and cumulative fundraising figures are company-supplied.” Neither report gives a precise post-money valuation.
What the Money Is For
According to Jiemian, the money will go mainly to video foundation-model research, training and inference infrastructure, product iteration and global expansion. The company said the full release of MAGI-2, a larger video-generation model built on a mixture-of-experts architecture, is coming soon (TechNode).
TechNode Global reported that Beijing-based Sand.ai was founded in 2024 and is led by Cao Yue, a former principal researcher at Microsoft Research Asia and a co-author of the Swin Transformer computer-vision architecture.
The Wider Video Race
TechNode Global placed the round after another large deal in China’s video-generation market: Kuaishou’s Kling AI unit completed a financing of roughly RMB19.8 billion in October, with Kuaishou keeping control of the business. In the US, Google cut the price of its Gemini Omni video model by 30% this week (our piece).
Our Numbers
$200 million of $300 million is 67% (our arithmetic), so this single round is twice what Sand.ai raised in its other two 2026 rounds combined. All three inputs (the round size, the ten-month total and the valuation above $1 billion) come from the company.
Connecting the Coverage
The raise lands in a week when AI funding marks moved in both directions: Groq was reported to be seeking a valuation 2.5 times its August mark (our piece), while listed AI cloud stocks fell 20.9% in four days (our piece). On Chinese model makers’ scale, Anthropic made 63 times the revenue of Zhipu and MiniMax (our piece).
The Structural Read
As we read the reports, the round’s size relative to the year is the clearest fact: one round is twice the other two combined, by our arithmetic.
As we read Wangsu’s role as TechNode Global reports it, the co-lead is also a planned infrastructure partner for inference and deployment, not only a financial backer.
As we read the timing, the company is raising before the full MAGI-2 release rather than after it; no benchmark results for that model have been published in these reports.
TechNode Global, 6 November 2026
“The valuation and cumulative fundraising figures are company-supplied.”
Three Implications
FOR READING AI FUNDING CLAIMS The unicorn status rests on a company-supplied valuation; neither report gives a precise post-money figure.
FOR CHINA’S VIDEO-MODEL MARKET TechNode Global sets the round beside Kling AI’s roughly RMB19.8 billion financing in October.
FOR THE MODEL ITSELF TechNode Global reports that Sand.ai’s website describes a MAGI-2 Preview as a 114-billion-parameter mixture-of-experts model; the company says the full release is coming soon.
The analysis says the supercycle “describes the intersection of three powerful forces: geopolitics, cultural transformation, and economic change.”
As we read it, this round mixes venture firms with investors that have state backing, as Jiemian reports, which puts capital and policy in the same cap table; the analysis does not discuss Sand.ai.
What Is Not Established
Neither the round’s exact valuation nor the size of each earlier 2026 round has been published. MAGI-2’s full release has no date, and we have not seen benchmark results for it. We have not seen Wangsu’s filing directly; its contents are as TechNode Global reported them.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Sand.ai says its $200 million Series A+ made it a unicorn and brought its 2026 fundraising to $300 million across three rounds. By our arithmetic the new round is 67% of that total. The valuation and fundraising figures are the company’s own.
A note on sourcing. The round, its investors, the ten-month total and the valuation claim come from Sand.ai as reported by Jiemian (in Chinese), TechNode and TechNode Global, linked in the text; the 67% share and the $100 million figure for the earlier rounds are our arithmetic. This is not investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
