Alibaba’s Wan3.0 targets business video at a fraction of Google Veo’s price

Alibaba has introduced Wan3.0, a video generation model that can create clips up to 30 seconds long in one run. Ana Maria Constantin reports for The Next Web that the launch follows Alibaba’s roughly $10.2 billion Hong Kong share placement, with all proceeds earmarked for AI investment.

Wan3.0 had been available in public beta through Alibaba Cloud’s Model Studio and Qwen Cloud since early August. The formal launch expands access, although the service remains staged. Users must apply for the beta, API access is still rolling out, and Alibaba plans a members-only consumer platform at wan.video.

The model accepts more than text prompts. Users can also provide images, audio, video, web pages, PDFs and PowerPoint files. Alibaba says this allows teams to turn structured material such as presentations and spreadsheets into videos. The company is positioning Wan3.0 for marketing, internal communications, social media and training simulations.

Longer clips and enterprise inputs

Wan3.0 doubles the 15-second limit of the earlier Wan2.7 model and supports output up to 1080p. Alibaba says it maintains characters, objects, layouts and motion across a clip, while producing synchronised facial expressions and multilingual voices. Those performance claims have not yet been independently benchmarked.

Pricing starts at $0.05 per second for 480p video, rises to $0.10 for 720p and reaches $0.20 for 1080p. At the top resolution, that equals $12 per minute. Alibaba’s price is below Google Veo 3.1’s standard rate of $0.40 per second, but some competing models cost even less.

The release also marks a change in Alibaba’s approach. Unlike earlier Wan models, Wan3.0’s weights are closed. Alibaba’s cloud and AI revenue grew 45% to 48.44 billion yuan in the June quarter, while capital expenditure climbed 75% to 67.68 billion yuan.