MidJourney debuts AI video tools to animate images into short clips: How does it work

AI Video & Visuals


Midjourney launched the first version of its video generation model powered by artificial intelligence, providing the ability for users to convert images into short animation clips. The ability to pay subscribers through both Midjourney's website and Discord Server indicates the platform's latest foray into creative generation tools.

In the current format, users can create 5-second videos from images that are generated within the platform or uploaded directly. A new “Animation” button appears after the image is rendered, and guides the user using a prompt-based animation workflow. The system applies the default motion sequence, but manual options are also available, allowing users to specify how images are moved.

Midjourney also enables video extensions, allowing users to extend animations in 4 seconds increments in total up to 21 seconds. Two movement modes are provided. Low and high, providing a different level of dynamism about how the subject or camera (or both) is animated.

The feature runs on GPU time models, starting at $10 a month with a high-speed GPU usage of around 3.3 hours, translated to almost 200 generations. However, video rendering is highly resource intensive, and it is estimated that 2 seconds of each video corresponds to the GPU cost of one image.

“This is just a stepping stone,” said David Holtz, founder of Mid Journey, in an official announcement. Holz noted that future versions of the tool could open the door to real-time open-world simulations and demonstrate broader ambitions for immersive media creation.

However, the rollout will take place at a time when legal pressure increases. Entertainment Giants Disney and Universal filed a lawsuit against the Mid Journey, accusing the company of promoting fraudulent regeneration of copyrighted material. The complaint labels the platform on “virtual vending machines” of derivative content, specifically citing the transition to video as a potential breach risk. The studio also questioned the legality of the training data used in the models.

As AI-generated content continues to advance, industry observers say the clash between innovation and copyright law has just begun.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *