In the last periods, the Chinese e-commerce giant Alibaba, who has done very valuable works for artificial intelligence, announced the new artificial intelligence model “Wan 2.1”. Wan 2.1, which can produce photos and images from the texts, seems to be mostly open source and unqualified.

Compared to the statements made by Alibaba, Wan 2.1 will have 4 different versions in the form of “T2V-1.3B”, “T2V-14B”, “I2V-14b-720p” and “I2V-14b-480p”. The most powerful of these was designed in the form that can process 14 billion parameters. But this artificial intelligence model has a valuable shortcoming. Even if you use the most advanced version, the highest resolution you can get is 720p.

Here is a video showing what Wan 2.1 offers

As you can see in the image above, Alibaba’s new artificial intelligence model can produce very realistic videos. The model has become so much that it is not possible to understand whether the image is real or straight. But let’s say that this is not surprising. Because, compared to Vbench, which is a test for artificial intelligence models that produce image, the most common model in this area is Wan 2.1.

If you want to quickly test Alibaba’s new artificial intelligence model, you can access Github through the link here and Huging Face through the link here.