Midjourney launched video—And it’s awesome!
The future is in motion!
Midjourney, long known to the AI art world for its stunning image generations, just pressed "play." Their brand-new video feature launched last week, and it’s not just a cool add-on—it might be the first step toward something much bigger: a fully interactive, AI-powered world in motion.
Welcome to Midjourney Video V1.
🎥 How it works: Image-to-Video
The new workflow is called "Image-to-Video," and it’s just what it sounds like. You generate a static image like you normally would, only now there’s a shiny new "Animate" button that lets you turn that image into a short, animated clip.
You can:
Use Automatic mode, which creates movement on its own
Use Manual mode, where you write a motion prompt describing how you want the image to move
There are also two settings for motion:
Low motion – subtle movements, great for ambient scenes (or... sometimes nothing happens at all)
High motion – everything moves, including the camera. Think drama. Think potential chaos. Think wow.
⏩ Extend it
Each video clip is about 5 seconds long, and you can extend the animation up to four times. That’s a total of 20 seconds of smooth, surreal footage if you keep hitting that extend button.
🔄 Upload & animate your own images
You’re not stuck with Midjourney-generated images, either. You can now drag in any image from your computer, set it as a "start frame," and tell Midjourney how it should animate. Yes, this means your old drawings, selfies, or even memes can be turned into incredible motion sequences.
For example, check out this photo of my Nana that I animated…
🤖 Real-time worlds are coming
According to Midjourney’s founder David Holz, this video feature is more than just a fun tool—it’s a major step toward real-time, open-world AI simulation.
"Imagine an AI system that generates imagery in real time. You can command it to move around in 3D space... and you can interact with everything."
They’re building the future piece by piece:
Image generation (done)
Video (in progress)
3D models (coming)
Real-time rendering (coming)
Full interactive AI environments (?!?)
💰 How much does it cost?
Each video job:
Costs around 8x more than an image job
Generates four 5-second videos
Comes out to about the same cost as a single upscale
Is 25x cheaper than what other AI video platforms charge
Monthly subscriptions range from $10 USD - $120 USD per month
A "relax mode" for video is also being tested for Pro subscribers and above.
🌟 Lisa’s demo: Astronaut in action
Want to see it in action? I created an image with the prompt:
"A hyper-realistic, cinematic image of an astronaut floating in zero gravity inside a space shuttle. No helmet, just a sleek iridescent and sequin spaceflight suit. The scene has vivid colors, high contrast. Shot on 35mm film."
…and hit “Animate”.
First, I tried automatic mode. Then I wrote my own motion prompt:
"A hyper-realistic, cinematic image of an astronaut floating in zero gravity inside a space shuttle. No helmet, just a sleek iridescent and sequin spaceflight suit. She drifts effortlessly, surrounded by glowing control panels, her hair is very long and drifts all around the shuttle.
The result? It was eerie. Cinematic. Like an AI-directed short film from the future.
🚀 Final thoughts: This is just the beginning
Midjourney Video V1 isn’t perfect—but it’s exciting. It gives artists, marketers, and dreamers a new tool to bring static visuals to life in seconds. And with their long-term roadmap pointing toward real-time simulation, this could be the dawn of AI worlds we can walk through.
So yeah, the image button just got a sibling. Say hello to Animate.
