Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

Alibaba launches Wan3.0 AI video model with 30-second generation

Alibaba has launched Wan3.0, its next-generation AI video generation model, capable of producing 30-second videos in a single generation. The release pushes the duration bar higher in a fiercely competitive video generation market and signals Alibaba's continued commitment to its foundation-model push.

Published
阿里发布Wan3.0视频生成模型:单次可生成30秒视频
Image source: alibaba.com

Alibaba has launched Wan3.0, its next-generation AI video generation model, with the headline capability of producing 30-second videos in a single generation.

The 30-second single-pass output stands out in a field where most models generate clips of a few seconds to around a dozen seconds, and few can produce a full half-minute of coherent footage in one go.

Longer single-pass generation demands stronger spatio-temporal consistency, motion control, and visual stability, and it directly shapes creators' productivity by bringing them closer to a one-shot workflow.

Alibaba's push into video generation is part of its broader multi-modal foundation model strategy, and Wan3.0 marks the latest major iteration of that line.

The video generation market is in the middle of an intense arms race, with players competing on duration, quality, and controllability; Wan3.0's bet on length is a clear strategic choice.

For creators, 30-second single-pass output could directly benefit short video, advertising, and pre-visualization work, potentially simplifying production pipelines.

What to watch next: how Wan3.0 is made available, its pricing, and independent evaluations of generation quality — all of which will determine whether it can hold its own among leading video models.

Why it matters

Wan3.0 resets the bar for single-pass video length, shifting the competitive focus in video generation from basic capability to duration, stability, and control.

AlibabaWan3.0Video Generation
Back to realtime news

Nearby Updates

All