Midjourney, an AI startup known for its surreal image generation tools, makes a bold leap into video. Recently, the company announced the V1, the much-anticipated video generation model that promises to bring static images to life. Midjourney is a big move as it throws the company into direct competition with other big hitters like Openai, Runway, Adobe, Google and more.
The V1 is designed as an image-to-video model, allowing users to convert either their uploaded photos or Midjourney's AI-generated images into short 5-second video clips. Like the sibling image model, the V1 is currently only accessible from Discord and is only available on the web at launch.
And it's not just Midjourney's visible video. In a blog post, CEO David Holz sets a rather ambitious goal for the company's AI, saying V1 is the next stepping stone to real-time “Open-World Simulations.” The company also revealed to the line plans to branch into 3D rendering and real-time generative models.
Midjourney's imaging tools have long appealed to artists and designers, but the company uses slightly different tacks from the video. Many of their rivals, such as Sora by Openai, Gen-4 by Runway, Firefly by Adobe by Google, and Veo 3, are chasing commercial filmmakers and studios with highly controllable AI tools. But Midjourney positions itself as a more creative playground for those looking for something a little more dreamy.
Midjourney V1 AI Video Generation Model: Pricing and Availability
Despite this, Midjourney is moving forward. However, video generation is not cheap. V1 still consumes eight times the credits per clip as Midjourney's image tool, so subscribers burn their monthly allowances much faster. At launch, basic subscribers who pay $10 per month (approx. $866) will have access to the V1, but unlimited video generation will be limited to $60 (approx. Rs. 5,200) pro and $120 (approx. 10,400) megaplans, and only to “relax” modes, which generate videos more slowly. However, the company said it will check this pricing structure in the coming weeks to collect feedback from users.
As for the tool itself, the V1 offers an incredible level of control. You can choose “automatic” mode where AI can generate your motion or “manual” mode, which accepts text prompts, to tell you exactly how you want the animation to work. Additionally, there are settings to adjust the intensity of movement. If a subtle shift is required, it is “low motion” or “high motion” for a more energetic effect. Clips last for 5 seconds by default, but can be extended up to 21 seconds in 4-second chunks.
Disney accuses Mid Journey of plagiarism
That said, Midjourney is in the video arena under the legal cloud. Just a week ago, Disney and Universal sued startups over image generation models, claiming they could create rogue versions of famous characters such as Darth Vader and Homer Simpson. As studios are nervous about AI tools that will replace human creatives, it is part of the growth of backlash across Hollywood, with AI companies facing questions about training data and copyright infringement.
Early examples of V1's output suggest that Midjourney is not aiming for hyperrealism, but rather sticking to its trademark. The initial response from users has been largely positive so far, but it's still too early to know how the V1 stacks up with more established players like Runway and Sora.
