Alibaba Launches Wan2.7 Image and Video Series to Support Content Creation

JAKARTA - Alibaba launched Wan2.7 Image and Wan2.7-Video, new Artificial Intelligence (AI) models for the multimedia ecosystem. These models are designed to simplify the content creation process.

Wan2.7 Image is designed to overcome the visual challenges of AI that have been inconsistent. This model allows the creation of professional images through deep personalization features that can set the structure of the face down to the smallest detail.

One of the flagship features of this model is precise color palette control. In addition, this model supports text rendering up to 3,000 tokens in 12 different languages for complex content needs.

Alibaba also provides a click-to-edit feature that will make it easier for users to accurately set image elements down to the pixel level. Users can also utilize up to nine reference images at once to speed up the creation of storyboards or architectural designs.

Next, Alibaba presents Wan2.7-Video which can turn text, images, and audio into 1080p quality video. With this model, users can set the movement of the camera as well as the emotional expression of the characters realistically.

This video model is able to maintain consistency of visual and voice identity for up to five different characters in a single long narrative. In this model, there is a video continuation feature to ensure a smooth and natural transition between scenes.

All of these latest models are integrated with the Qwen App and are available through Alibaba Cloud Model Studio for large-scale use. These two models are believed to encourage the creativity of content industry players.