Midjourney fans are in for a treat. The company has just released its first official video generation model, V1. This new tool lets users create short videos from images, either uploaded or generated by Midjourney. The process is simple: upload an image, and V1 will create a 5-second video clip. Currently, this feature is available on Discord and can be accessed through the web interface.
Key Features
The V1 model comes with some exciting features that make video creation a breeze.
- Users can choose between automatic animation mode, which creates a random motion, or manual mode, where they can specify the movement using text prompts.
- The model also allows for camera and object motion adjustments, giving users more control over their video output. They can opt for “low motion” or “high motion” settings to achieve the desired effect.
- Another notable feature is the ability to extend the video length. While the initial clip is 5 seconds long, users can add up to 4 seconds at a time, with a maximum of 4 extensions, resulting in a total video length of 21 seconds.
Midjourney’s CEO, David Holz, says the company’s ultimate goal is to create AI models that can simulate real-world environments in real-time. The V1 model is just the first step towards achieving this goal. The company plans to develop AI models for generating 3D images and real-time simulations in the future.
Pricing
Using the V1 model comes at a cost. Video generation is eight times more expensive than image creation, which means users will consume their monthly quotas faster. The Basic plan, priced at $10 per month, is the entry point for trying out V1. Pro and Mega plan subscribers, who pay $60 and $120 per month, respectively, get unlimited video generation in ‘Relax’ mode, which processes requests at a slower pace.
Copyright Concerns
Midjourney has faced lawsuits over training its AI using images from Disney without permission. While the company’s focus on creating high-quality visuals makes it a favorite among creatives, the issue of using copyrighted material without consent remains a concern. Other AI companies are also grappling with this problem, and it remains to be seen how the industry will address it as AI technology advances.