Alibaba has released Wan 3.0, the next version of its video generation artificial intelligence model. A few days ago, the IT giant raised around $10.2 billion through a share issue on the Hong Kong Stock Exchange.
Image source: Alibaba Cloud
This service has been tested internally on Alibaba Cloud and Qwen cloud platforms in early August, and Wan3.0 models are currently fully online. The updated neural network can produce 1080p resolution video with a duration of up to 30 seconds, twice the capability of the previous version Wan2.7. According to the developers, the AI model preserves character details, spatial layout, and animated graphics throughout the film, and also allows you to add voiceovers in different languages.
The most notable feature of Wan3.0 is what types of input data the neural network can process. In addition to text, images, audio and video, the service interacts with web pages and documents, including PDF and PowerPoint files. This allows users to upload a presentation or spreadsheet and receive a video that reflects the structure of the source material.
As a result, Alibaba’s new products are aimed more at marketing departments and corporate communications teams rather than media content creators. Meanwhile, developers believe Wan 3.0 will be useful in a wider range of use cases, including creating videos for social networks and simulation training in autonomous transportation and robotics. Essentially, neural networks are a business tool disguised as a creative tool.
Pricing follows this logic. Wan3.0 480p videos cost $0.05 per second, 720p videos cost $0.1 per second, and 1080p videos cost $0.2 per second. Therefore, users must pay $12 per minute when creating movies at the highest resolution available. In comparison, the standard Google Veo 3.1 plan charges $0.40 per second for video.
If you find an error, select it with your mouse and press CTRL+ENTER.
