Midjourney has announced their new video creation model
For months, fans have been anticipating the arrival of a text2video feature from this startup, and finally, it has happened.
In the spirit of Midjourney, the main goal during development was to create beautiful and detailed images based on text prompts. For example, now you can adjust the level of motion dynamics in the video or control its style. In addition to generating videos from text, there is also the ability to animate existing images.
Another interesting point is that the initial video length is only 5 seconds, but it can easily be extended up to 20 seconds if desired. And if you’re unsure about what exactly you want to create, there is a fun random generation mode: the model will select something unexpected and interesting for you without any prompt input.
Overall, the resulting videos are quite atmospheric and vibey, especially if you put effort into crafting your prompt. You can try it yourself here — midjourney.com/home.