Midjourney launches first AI video tool to animate images

Midjourney has launched its first public video generation tool, allowing users to create animated clips from still images.

The new feature, announced on Midjourney’s website, enables the generation of five-second videos from either uploaded images or those created within the Midjourney platform.

Once an image is selected, an “animate” button appears, offering a standard motion option or a manual setting for customising movement.

The animation can be extended in four-second increments, with a maximum length of 21 seconds. Motion settings allow users to determine whether the subject, the camera, or both are animated.

The feature is currently limited to subscribers, with plans starting at $10 (USD) per month.

Midjourney says the cost of video generation is approximately eight times higher than that of image rendering, equating to roughly one image’s worth of cost per second of video.

The launch comes amid an ongoing copyright lawsuit filed by Disney and Universal.

The two entertainment companies argue that Midjourney’s training process may have included unauthorised use of copyrighted works. Legal filings specifically flagged the video tool as a concern, with plaintiffs describing the company’s output as a “virtual vending machine” for unlicensed content.

Midjourney founder David Holz described the release as “a stepping stone” in the company’s development of real-time, open-world simulation models.

The startup is one of several firms racing to develop AI-powered video tools, with competitors such as Google, OpenAI, and Meta also releasing systems that generate video content from text prompts.

Social media users have been sharing their first reactions, including early videos, with many hailing it as a ‘game-changer’.

While Midjourney’s current offering focuses on animating static images, the company has indicated that expanded capabilities will be introduced in future updates.