Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]

reuters+1reutersthenextwebAlibaba officially launched Wan3.0, the latest version of its AI video generation model, on Monday, capable of producing 30-second videos from documents, spreadsheets, slides, and web pages. The release comes one day after the company announced a roughly $10.2 billion Hong Kong share placement — the largest-ever primary follow-on offering by a Hong Kong-listed company — with proceeds earmarked entirely for AI investment.thenextweb+1
The model doubles the output length of its predecessor Wan2.7, generating clips of up to 30 seconds in a single pass at resolutions up to 1080p. Alibaba says it maintains character detail, spatial layout, and motion consistency across the full duration, with synchronized facial micro-expressions and multilingual voice output.thenextweb
Its most distinctive capability is the range of inputs it accepts. Beyond text, images, audio, and video, Wan3.0 ingests DOC, XLS, PPT, PDF, and Markdown files, allowing users to hand it a slide deck or spreadsheet and receive a structured video in return. That positions the tool squarely at corporate communications and marketing teams, though Alibaba also lists short dramas, social content, and simulation training for autonomous vehicles among its use cases.wtvbam+2
Pricing runs at $0.05 per second at 480p, $0.10 at 720p, and $0.20 at 1080p — working out to $12 per minute at the top tier, well below the $0.40 per second standard tier of Google's Alphabet Inc. Veo 3.1. A temporary 30% discount on API pricing is available from August 24 through September 23.technode+1
Wan3.0 had been running in public beta since August 6 through Alibaba Cloud's Model Studio and its Qwen Cloud platform, so Monday's announcement widens access rather than marking a first appearance. Full API availability is described as rolling out, and a consumer-facing site at wan.video is planned as a members-only platform, suggesting Alibaba is metering demand while managing compute costs.wtvbam+1
Notably, the model's weights are closed. Alibaba built much of its AI reputation by open-sourcing earlier versions of the Wan series, with Wan2.2 in July 2025 the last video model released openly — a shift that accompanies its decision to begin charging heavy users of its open Qwen language models.thenextweb
The commercial logic behind the launch showed up in Alibaba's most recent earnings. Cloud and AI revenue grew 45% year-over-year to 48.44 billion yuan in the June quarter, while AI model services alone now carry more than 16 billion yuan in annualized recurring revenue. The cost was equally visible: quarterly net profit fell 75% as capital expenditure rose 75% to 67.68 billion yuan. Alibaba committed 380 billion yuan to AI infrastructure over three years in early 2025 and has reportedly been weighing an increase to 480 billion yuan.wtvbam+1
Whether Wan3.0 justifies that spending remains an open question. The model has not yet been independently benchmarked, making its quality claims company-issued for now.thenextweb