Midjourney, one of the hottest startups of the AI image generation, announced on Wednesday starting his long -awaited model of AI, V1 video generation.
V1 is a picture model for a film in which users can send a picture-like to make a picture generated by one of the other Midjourney-a V1 models will produce a set of 4 five seconds based on it. Like the Midjourney painting models, V1 is only available via Discord and is only available on the web in launching.
The introduction of V1 introduces Midjourney in competition with AI video generation models from other firms, resembling Sorai Sorai, Runway Gen 4, Adobe’s Firefly and Google Veo 3. While many firms focus on developing controlled AI video models for use in industrial settings, Midjourney at all times distinguished their characteristic models AI image.
The company claims that it has larger goals for its AI video models than generating B-ROLL for movies or promoting in the promoting industry. IN blog Post, David Holz, general director of Midjourney, says that his AI video model is the next step of the company towards the final destination, creating AI models “able to simulate open world in real time.”
After video models, AI Midjourney claims that he plans to develop AI models for the production of 3D rendering, in addition to AI models in real time.
The launch of the V1 Midjourney model appears only a week after the startup was sued by two of the most famous film studies in Hollywood: Disney and Universal. The lawsuit claims that the paintings created by the models of paintings AI Midjourney show characters protected by copyright resembling Homer Simpson and Darth Vader.
Hollywood Studios tried to confront the growing popularity of the image of AI image and video generating models resembling these midjourney. There is a growing concern that these AI tools can replace or devaluate the work of creation in their fields, and several media firms claimed that these products are trained in their very own law -protected works.
While Midjourney tried to tell apart himself from other starts of the AI and Video image – more focused on creativity than immediate industrial applications – the startup cannot avoid these accusations.
At the starting, Midjourney claims that he’ll load 8x more for video generating than typical image generation, which implies that subscribers will run out of monthly designated generations much faster when creating movies than images.
During the premiere, the most cost-effective strategy to try V1 is the subscription to the basic Midjourney plan value USD 10 monthly. Subscribers of the Pro Midjourney plan in the amount of USD 60 monthly and a mega plan of 120 USD monthly will have unlimited video generations in the slower “Relax” mode. Over the next month, Midjourney claims that the price of video models will again assess the price again.
V1 is equipped with several custom settings that allow users to manage the results of the video model.
Users can select the automatic setting of the animation so that it randomly move or can select a manual setting that permits users to explain a specific animation in the text that they need to add to their film. Users may switch the amount of camera and the subject, choosing “low traffic” or “high traffic” in the settings.
While movies generated with V1 have only five seconds, users can select them by 4 seconds to 4 times, which implies that V1 movies may be as much as 21 seconds.
Like the AI Midjourney image models, the early versions of V1 movies look a bit unearthly, not hyperrealistic. The initial response to V1 was positive, although it is still not clear how well it matches other leading AI video models that have been on the market for months or even years.