The creators of the renowned image-generating platform Midjourney have just taken a bold step forward, launching their highly anticipated V1 Video Model. This release marks the company’s first foray into motion-based AI and signals a future where immersive, real-time simulations could become a reality.
For years, Midjourney has been synonymous with high-quality AI-generated visuals. Now, it’s laying the groundwork for an even more ambitious vision — one that includes fully interactive 3D environments and dynamic video generation. According to Midjourney’s founder, David Holz, this new direction is essential for building AI capable of simulating living, moving worlds in real time.
A New Chapter Begins: Introducing “Image-to-Video”
The newly released tool, dubbed Image-to-Video, enables users to animate the images they already create in Midjourney with just a click. The feature is designed to be user-friendly and fun, offering both automatic and manual animation options.
- Automatic Mode: Creates movement based on AI-generated motion prompts.
- Manual Mode: Allows users to input detailed motion instructions for customized animations.
Additionally, users can choose between “low motion” for subtle, atmospheric scenes and “high motion” for more dynamic camera and subject activity. However, each mode comes with trade-offs — low motion might result in minimal movement, while high motion can occasionally cause visual inconsistencies.
Extend, Upload, Animate
The video clips generated through Midjourney’s tool are around 5 seconds long, and users can extend them up to four times for a total of 20 seconds. Another exciting feature is the ability to animate external images — simply upload an image, label it as a starting frame, and input a motion prompt to bring it to life.
Balancing Access, Cost, and Sustainability
The company is starting with a web-only rollout, with video tasks priced at approximately eight times the cost of standard image generation. Each video job creates four short clips — a cost that Midjourney says is on par with the current cost of upscaling images and dramatically cheaper than existing video generation tools on the market.
Pro-tier users can also expect a “relax mode” for videos, similar to the one offered for image generation, allowing for more flexible use at a lower priority.
What’s Next?
While this is just an initial step, the Midjourney team sees video generation as a key building block for even more complex systems — eventually combining image, video, 3D, and real-time models into a single cohesive simulation platform.
