Midjourney, known for its AI still image generation technology, recently announced the launch of its first AI video generation tool, thus expanding into the field of video creation.
Midjourney's first AI video generation tool can turn still images into short videos up to 21 seconds long. The source material can be images generated by Midjourney or uploaded by the user. This tool is currently only available as a web version and requires a $10 monthly subscription.
Introducing our V1 Video Model. It's fun, easy, and beautiful. Available at 10$/month, it's the first video model for *everyone* and it's available now. pic.twitter.com/iBm0KAN8uy
— Midjourney (@midjourney) June 18, 2025
Starting from images, injecting "dynamic imagination" into creation
The video generation tool released this time is still an early test version. The default length of a single video is 5 seconds, but users can choose to extend it up to 4 times, 4 seconds each time, so that the generated video can be up to 21 seconds long.
While not as fully capable of generating video (or even audio) from text as OpenAI's Sora or Google's recently launched Veo 3, Midjourney emphasizes that its tool still focuses on "starting from creating images and bringing them to life," emphasizing the importance of maintaining an artistically expressive design. Therefore, compared to video generation tools like Adobe's Firefly Video, Google Flow, or Runway, Midjourney's video generation tool extends beyond static aesthetics to dynamic storytelling.
Copyright disputes continue to rage, and creators' concerns remain
However, Midjourney has previously faced lawsuits from companies like Disney and NBCUniversal, alleging that it illegally used copyrighted content to train models. This isn't the first time Midjourney has faced similar controversy, and its move to provide video generation tools may further challenge the boundaries of content creation rights.
In response to external criticism, Midjourney simply urged users to "please use these technologies responsibly," without further detailing the source of training data or countermeasures. Given the vulnerability of AI models to misuse to generate false content, deepfakes, or infringe upon the creative work of others, the industry is calling on relevant companies to strengthen transparency and licensing mechanisms.
As you know, our focus for the past few years has been images. What you might not know, is that we believe the inevitable destination of this technology are models capable of real-time open-world simulations.
What's that? Basically; imagine an AI system that generates imagery in…
— Midjourney (@midjourney) June 18, 2025
