Midjourney releases the first new AI image model for almost one year


Midjourney, one of the earliest services revolving the AI ​​web development, releasing the first new image model of AI in almost one year.

Called dubbed V7The model, starting to roll the alpha at midnight on the east Friday, it took a week after OPUII a new IMAGE generator in Chatgpt that easily riding Viral For the ability to make pictures of Ghmibri-style. Midjourney’s is not a ghatrized model – at least not officially – but however aesthetically pleasing acts, at least in this reporter’s reporter.

To use it, you need to rate about 200 images to build a “PersonalIIS” profile in Midjourney “if you have no longer. This Profile is the first midjourney with a personalization to be defaulted by default.

If you have finished that task, you are greeted by a togglable V7 on the Midjourney website and, if you are membered at the discord server in Midjourney. In the web app, you can select the model immediately from the drop-down menu next to the “version” label.

Midjourney V7
Credits in the image:Midjourney

Midjourney CEO David Holz characterized by V7 as a “different different architecture” in a post to x.

“V7 (…) much better than the texts that prompted,” Holz continued a contention of conflict. “(I) advocate the prompts look fantastic, the quality of the image is higher than beautiful textures, and bodies, hands on all kinds of all situations.”

V7 is available in two flavors, turbo and relaxing – the first more expensive run – and power of a new 10x-drafting tool costs standard mode. Draft images are lower quality than standard-mode images, but can be enhanced and reacted to a mouse click.

Many Midjourney capabilities are not yet available for V7, according to Holz, including dismantling of image and withdrawal. That will come to almost come, he says – probably when within two months.

“It’s a new model with extraordinary strength and maybe some weaknesses” Holz wrote to the friction. “(W) want to learn from you how good it is good and bad for, but definitely remember it will need different styles of prompting.”

Midjourney V7
My prompt: “A dragon with a spiky tail.”Credits in the image:Midjourney

In my short test, V7 follows reasonable healing of the promptings I give. True, I don’t have time to certainly Put the model through the ringer.

Midjourney V7
My Prompt: “A trait style of GRAB.”Credits in the image:Midjourney

Midjourney is an unusual operation. Beginning at 2022 by Holz, hiding the PC peripheral company to jump into motion, it doesn’t get dime outside money.

In late 2023, Midjourney IS reported Looking forward to bring around $ 200 million income. Recently, San Francisco-based company says this Establish a hardware team To work with unknown projects, and it continues to practice previously informed models for video generation and 3D.

many Laws Accusing midjourney to violate the rights of millions of artists by training AI items in images without permission of images’ inferences.





Source link

  • Related Posts

    Meta exec denies artificial company enlarged llama benchmark marks

    A Meta Exec on Monday a gossip a gossip with the company has been trained new AI models to present the weaknesses of benchmarks while hiding weaknesses in models. The…

    Tom Cruise Gear Up To Save Us From Ai In New New Mission: Impossible – The final counting trailer to count

    In the last mission: Impossible film, Dead CountLong-term franchise has been officially entered in the territory of Science Fiction by making its intelligent villain. We saw ethan of ethan Tom…

    Leave a Reply

    Your email address will not be published. Required fields are marked *