Midjourney Video Model V1 | Rapid Realistic Video Generation

Midjourney announced the launch of a new video model — V1

It is now available to all subscribers with a minimum payment of $10. Currently, the resolution is limited to 480p, and the generation time is just five seconds. I think this was intentionally done for the initial launch to prevent server overheating, while users can continue to generate individual segments. For each prompt, the system immediately produces four video clips, just like before with images in the standard Midjourney — all quick and convenient.

The videos look very neat: without noticeable artifacts, but the level of realism still leaves something to be desired. The company has been working on video creation for almost two years. Why so long? Generating moving images is a very resource-intensive process, and Midjourney relies solely on its own profits and has never attracted external funding.

It’s been a long journey to this innovation — it’s interesting to see how much effort has been invested in development.

Page view 25.04 11:29 Page view /ai-blog/vibevoice-update-real-time-high-quality-text-to-speech-microsoft/ 25.04 11:26 Page view /ai-blog/top-open-api-ai-services-fast-reliable-cost-effective-ai-api-experts/ 25.04 11:19 Page view 25.04 11:18 Page view /ai-blog/ai-in-schools-bridging-the-digital-divide-for-future-skills/ 25.04 11:14 Page view 25.04 11:13 Page view /ai-blog/ai-in-business-management-lessons-from-anthropics-experiment/ 25.04 11:08 Page view /ai-blog/nvidia-new-memory-approach-faster-large-language-models-nvidia/ 25.04 11:06 Page view 25.04 11:02 Page view 25.04 10:56