Midjourney has launched Version 1 of its long-awaited video model, enabling users to animate still images using AI-generated motion. Founder David Holz presented a vision for real-time open-world AI simulations, which will integrate static image generation, motion, 3D navigation, and real-time responsiveness over the next year as modular capabilities. The new video feature operates with an “Animate” button, offering automatic movement based on context and custom motion prompts for user-directed animations. Users can choose between “low motion” and “high motion” settings for varied animation intensity and can upload external images as start frames. Videos can extend up to 20 seconds, with pricing reflecting the increased computational load, costing about eight times more than image jobs, yet remaining 25-fold cheaper than other commercial options. Though the video feature is currently web-exclusive, demand may exceed capacity, prompting potential fluctuations in pricing and availability as Midjourney adjusts to early usage patterns.
Source link
