AI image generator Midjourney has rolled out an AI image-to-video generator for the first time, bringing it into competition with the likes of OpenAI’s Sora, Google’s Veo 3, and Adobe’s Firefly.
The tool, called V1, is currently only available to desktop users via Midjourney’s Discord app, and you’ll need to sign up for the company’s $10-a-month subscription at a minimum to start using the service.
Users can press “Animate” to make their Midjourney-generated images move. V1 will then produce a set of four five-second videos based on whatever they input.
There are two animation settings you can pick from: “automatic” and “manual.” With the “automatic” setting, the tool creates what the startup calls a “motion prompt” and “just makes things move.” But if you’re after a bit more creative control, the “manual” animation button lets users describe how they want things to move and how the scene should develop.
(Credit: Midjourney)
V1 also offers two styles: high motion and low motion. In low motion, the camera stays mostly static while the subject moves slowly or deliberately. In high motion, both the subject and the camera move—though Midjourney admits that “all this motion can sometimes lead to wonky mistakes.”
In addition, users can animate images uploaded from outside of Midjourney. You’ll need to drag the image you want to animate into the prompt bar and mark it as a “start frame,” then type a motion prompt to describe how you want it to move. Once you have a video you’d like to hold on to, you can “extend” it, roughly four seconds at a time, up to four times total.
But turning your creative aspirations to video won’t come cheap: using Midjourney to generate video will cost eight times more than conventional image generation, meaning you burn through your monthly credits much faster than normal.
Recommended by Our Editors
There’s also no guarantee of what the tool will ultimately end up costing at this early stage. Midjourney noted that the cost of running these models is “hard to predict,” and that it monitors how people use the service before adjusting pricing to ensure it’s running “a sustainable business.”
The new features come as Midjourney has plenty on its plate to deal with beyond product design. Last week, Universal and Disney sued the Bay Area start-up, claiming its business is “a bottomless pit of plagiarism,” as a result of drawing from many of the studios’ iconic productions for its imagery.
But lawsuits aren’t stopping the startup from making highly ambitious pronouncements about the future of its tech. “We believe the inevitable destination of this technology is models capable of real-time open-world simulations,” said a Midjourney spokesperson as part of the announcement.
5 Ways to Get More Out of Your ChatGPT Conversations
Get Our Best Stories!
Your Daily Dose of Our Top Tech News
By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy Policy.
Thanks for signing up!
Your subscription has been confirmed. Keep an eye on your inbox!
About Will McCurdy
Contributor
