On December 16, 2025, Alibaba made another big move as the new generation of Wanxiang 2.6 series models were officially launched! This model has been comprehensively upgraded for professional film and television production and image creation scenes, and is also praised as "the most versatile video generation model in the world". Moreover, it has been simultaneously unveiled on the official websites of Alibaba Cloud Bailian and Wanxiang, waiting for everyone to experience it!

Upgrade content of Wanxiang 2.6 model:

New "Character" function, set any person or object as a character to perform your A-list blockbuster

Video intelligent multi lens, fully upgraded narrative ability, smooth switching between multiple lenses, up to 15 seconds

More beautiful and artistic image generation, with comprehensive upgrades in consistency, controllability, and expressiveness

There are many highlights in the core upgrade, and the breakthrough in duration has become the focus

The upgrade of Wanxiang 2.6 has further improved in terms of image quality, sound effects, and command compliance. Even more impressive, it has upgraded the duration of a single video to the highest in China, 15 seconds. You should know that many video generation models in the past had time limitations, but the breakthrough of Wanxiang 2.6 undoubtedly provides creators with a broader space for development.

The capabilities of this model family are also very strong, supporting more than 10 visual creation abilities such as text to image editing, text to video, image to video, voice to video, action generation, role-playing, and general video editing. So many functions in one, it's like a "treasure box" for video creation.

1. Role playing function: pioneering in China, full sensory transfer

One of the biggest highlights of the Wanxiang 2.6 series is that it is the first video model in China to support "role-playing" functionality. How magical is this feature? It can refer to the appearance and tone of characters in the input video, and generate single person, multi person, and person object co production videos according to the prompt words. Just like, if you want a celebrity and a virtual character to shoot a video together, Wanxiang 2.6 can help you achieve it.

In terms of model structure, Tongyi Wanxiang integrates multiple innovative technologies. It will perform multimodal joint modeling and learning on reference videos, extracting subject emotions, postures, visual features, as well as acoustic features such as timbre and speech rate.

This ensures consistency and transfer across all senses and dimensions, making the generated video look natural and realistic. Isn't this a major innovation breakthrough in the field of video creation? Compared to previous models that could only generate videos simply, the role-playing function of Wanxiang 2.6 is simply too powerful.

2. Professional level storyboard control: simple prompts, coherent narrative

In addition to role-playing function, Wanxiang 2.6 also adds storyboard control function. This feature is so practical, it can convert users' simple prompt words into multi shot scripts, generating coherent narrative videos containing multiple shots. Previously, to create a video with multiple shots, it required professional screenwriters and directors to design storyboards, which was a tedious and complex process.

But now, with Wanxiang 2.6, through high-level semantic understanding, it can construct professional level multi shot paragraphs with complete storylines and narrative tension. Moreover, during the smooth switching of the lens, it can maintain a high degree of consistency in the core subject, scene layout, and environmental atmosphere. It's like equipping each creator with a professional director assistant, making video creation easier.

Empowering film and television level creation, anyone can become a director

The role-playing and storyboard control functions of Wanxiang 2.6 greatly meet the needs of professional film and television level scenes. For example, an ordinary user can upload a personal video and input science fiction suspense style prompts. With Wanxiang 2.6, it only takes a few minutes to complete storyboarding design, character interpretation, screen dubbing, and other tasks, generating a short film with a complete narrative and cinematic quality. Isn't this just helping ordinary users fulfill their dream of becoming movie protagonists?

For professional scenarios such as advertising design and short drama production, Wanxiang 2.6 also performs well. By inputting continuous prompt words, the model can generate a complete narrative short film. This means that in the future, everyone can become a director and create their own wonderful video works. Imagine in the future, we may see more excellent videos created by ordinary people, how interesting they would be!

It is worth mentioning that Alibaba was the first to release the Wanxiang 2.5 video generation model for audio and image synchronization in China in September this year. On the authoritative large model evaluation set LMArena, Wanxiang's video generation model ranked first in China. The release of version 2.6 undoubtedly further consolidates its leading position in the domestic video generation field.

Diverse experience modes and rich gameplay are about to be launched

Starting today, everyone can directly experience the powerful features of Wanxiang 2.6 on the Wanxiang official website, and enterprise users can also call APIs through Alibaba Cloud Bailian.

Moreover, it is reported that the Qianwen APP will also launch the model in the near future and provide richer gameplay options. At that time, we will be able to experience the charm of video creation brought by Wanxiang 2.6 on more platforms.

The official launch of Wanxiang 2.6 has brought new vitality and opportunities to the field of video creation. It not only enables professional creators to complete their works more efficiently, but also gives ordinary users the opportunity to realize their creativity.

The editor believes that in the future, Wanxiang 2.6 will play a greater role in the field of video creation, creating more exciting works.