Midjourney Launches V1: A Bold Step into AI Video Generation
Midjourney, one of the most popular AI image generation startups, announced on Wednesday the launch of its much-anticipated AI video generation model, V1.
V1 is an image-to-video model that allows users to upload an image or use one generated by another Midjourney model and receive a set of four five-second videos based on it. Much like Midjourney’s image models, V1 is only accessible through Discord, and it is currently only available on the web at launch.
The release of V1 puts Midjourney in direct competition with other AI video generation tools, including OpenAI’s Sora, Runway’s Gen-4, Adobe’s Firefly, and Google’s Veo 3. While many companies are focused on building controllable AI video models for commercial use, Midjourney has consistently stood out with its distinct, creativity-focused AI image models.
According to a blog post by Midjourney CEO David Holz, the company has broader ambitions for its AI video models beyond simply generating B-roll for Hollywood or ad agencies. Holz describes the launch of V1 as a step toward their ultimate goal: creating AI systems “capable of real-time open-world simulations.”
Following the development of its video models, Midjourney plans to expand into 3D rendering and real-time AI models.
Interestingly, the launch of V1 comes just a week after Midjourney was sued by two major Hollywood studios: Disney and Universal. The lawsuit claims that images generated by Midjourney’s models include copyrighted characters like Homer Simpson and Darth Vader.
Hollywood studios have struggled to keep up with the rapid rise of AI-generated content. There’s growing concern that tools like Midjourney’s could devalue or replace the work of human creatives. Several media companies have also alleged that these AI systems are trained on copyrighted materials without consent.
Despite trying to differentiate itself from other AI startups emphasizing creativity over commercial application Midjourney hasn’t been able to avoid these legal challenges.
As for pricing, video generations will cost eight times more than image generations, meaning subscribers may exhaust their monthly quota much faster when generating videos.
At launch, the most affordable way to try V1 is through Midjourney’s $10-per-month Basic Plan. Those on the $60-per-month Pro Plan and $120-per-month Mega Plan will enjoy unlimited video generations in the platform’s slower “Relax” mode. Midjourney says it will reassess its video pricing model over the next month.
V1 also introduces several customizable settings for controlling video output.
Users can opt for automatic animation, which randomly animates the image, or select manual mode to describe a specific animation in text. The amount of motion in the camera or subject can also be controlled using “low motion” or “high motion” settings.
Though V1 videos are initially only five seconds long, users can extend them by four seconds up to four times allowing for a maximum video length of 21 seconds.
Similar to its image models, early demos of V1’s output appear otherworldly and artistic, rather than photorealistic. Initial feedback has been positive, though it’s still unclear how V1 will stack up against established players that have had more time on the market.
