Alibaba's Wan3.0 Officially Launches, Stock Falls ~3% in Pre-Market Trading
Nashnova编辑部
Alibaba on August 25 officially launched Wan3.0, an AI video-generation model that turns documents, spreadsheets, and web pages into videos up to 30 seconds long; U.S.-listed shares fell about 3% pre-market, as investors wait for proof that the company's multi-billion-dollar AI bet can generate real revenue.
What can Wan3.0 actually do?
Users upload documents, spreadsheets, slides, or web pages, and the model auto-generates videos up to 30 seconds long.
This means → the target is not "type a prompt, get a clip." It is converting a company's existing files directly into video content.
Since a public beta on August 6, the model has been used in short dramas, ad campaigns, tourism promotion, and music videos. The official launch marks a shift from testing to full commercial deployment.
Why does the timing matter?
Wan3.0 arrives right after Alibaba completed a $10.3 billion Hong Kong share placement — 710 million shares at HK$112.70 each.
The proceeds are earmarked for full-stack AI infrastructure, including expansion and upgrades.
In plain terms = Alibaba just raised a massive war chest labeled "AI." Wan3.0 is the first visible checkpoint for whether that spending can deliver commercial returns.
What does the Apple tie-up mean?
Apple recently published a guide for mainland China Mac users on connecting Alibaba's AI services to Siri and Writing Tools.
China's internet regulator has approved Apple Intelligence for mainland iPhones; Alibaba's Tongyi Qianwen (Qwen) will integrate with Apple Intelligence across iPhone, iPad, Mac, and Vision Pro.
This means → Alibaba's AI gains a distribution channel inside Apple's ecosystem — users access the model without downloading a separate Alibaba app.
Alibaba is not exclusive, though — Baidu is also partnering with Apple on similar features. The competitive landscape remains open.
Why did the stock fall anyway?
On the day of the announcement, Alibaba's U.S.-listed shares dropped about 3% pre-market.
This reflects the market's current stance on AI product launches: show me the revenue first, then I'll re-rate the stock.
In plain terms = investors are not asking "Is the model any good?" They are asking "Can short dramas and ads actually scale into meaningful revenue?" — until that answer is clear, the share price is digesting the post-placement overhang more than pricing the product.
Content is for reference only, not financial advice.