Midjourney Launches First AI Video Generation Model, Exploring “World Models”

Despite being sued by Disney, MidJourney has launched its first AI video generation model, V1, with the ultimate goal of creating a “world model.” The system aims to integrate static image generation, animation creation, 3D spatial navigation, and real-time rendering into a unified platform. This will allow users to freely explore a simulated environment, where visuals, behaviors, and user interactions can be dynamically realized, much like immersive video games or VR experiences. The envisioned future scenario is simple: a user could input a phrase such as "take me to an underwater city during a sunset," and the system would generate an explorable world with continuously evolving visual outcomes, potentially even including dynamic soundscapes. This vision positions V1 as a stepping stone, a crucial piece of technology in the foundation that MidJourney is building. Currently, however, V1 is far from perfect. Compared to competitive tools that have been operating in the image generation space for months or even years, V1 still shows significant gaps in functionality and technological maturity. Nevertheless, MidJourney's strong brand influence in the image generation domain and its user base suggest that V1’s introduction opens up new growth opportunities for the company. More importantly, it marks a rapid shift in AI content creation tools from single-media formats to multi-media integration, signaling potential profound changes in the way content is created and consumed. The launch of V1 is an important milestone in MidJourney’s journey toward a more complex and holistic AI system. Each technological iteration brings the company closer to its ultimate vision. While V1 has its limitations, its introduction demonstrates MidJourney's commitment to pushing the boundaries of AI-generated content and exploring new ways to create and interact with digital environments. For reference: 1. https://www.midjourney.com/updates/introducing-our-v1-video-model 2. https://techcrunch.com/2025/06/18/midjourney-launches-its-first-ai-video-generation-model-v1/ Editor/Arranger: He Aolong