Midjourney, one of many first AI-based picture turbines on the net, has launched its first new AI picture mannequin in almost a yr.
Dubbed V7, the mannequin started rolling out in alpha round midnight EST on Thursday, comes every week after OpenAI debuted a brand new picture generator in ChatGPT that shortly went viral for its skill to create Ghibli-style pictures.
Midjourney’s shouldn’t be a Ghibli-optimized mannequin, although — at the very least not formally — however it nonetheless can generate aesthetically pleasing works, at the very least to this reporter’s dilettante eye.
We’re now starting the alpha-test part of our new V7 picture Mannequin. It’s our smartest, most stunning, most coherent mannequin but. Give it a shot and anticipate updates each week or two for the following two months. pic.twitter.com/Ogqt0fgiY7
— Midjourney (@midjourney) April 4, 2025
To make use of it, you’ll first need to fee round 200 pictures to construct a Midjourney “personalization” profile, for those who haven’t already. This profile tunes the mannequin to your particular person visible preferences; V7 is Midjourney’s first mannequin to have personalization switched on by default.
When you’ve executed that, you’ll be capable to flip V7 on or off on Midjourney’s web site and, for those who’re a member of Midjourney’s Discord server, on its Discord chatbot. Within the internet app, you’ll be able to shortly choose the mannequin from the drop-down menu subsequent to the “model” label.
Midjourney CEO David Holz described V7 as a “completely totally different structure” in a put up on X.
“V7 is […] a lot smarter with textual content prompts,” Holz continued in an announcement on Discord. “[I]mage prompts look improbable, picture high quality is noticeably increased with stunning textures, and our bodies, palms, and objects of every kind have considerably higher coherence on all particulars.”
V7 is out there in two flavors, Turbo (costlier to run) and Loosen up, and powers a brand new software referred to as Draft Mode that renders pictures at 10x the pace and half the price of the usual mode. Draft pictures are of decrease high quality than standard-mode pictures, however they are often enhanced and re-rendered with a click on.
Numerous customary Midjourney options aren’t obtainable but for V7, based on Holz, together with picture upscaling and retexturing. These will arrive within the close to future, he mentioned, presumably inside two months.
“That is a wholly new mannequin with distinctive strengths and possibly a number of weaknesses” Holz wrote on Discord. “[W]e need to study from you what it’s good and dangerous at, however undoubtedly be mindful it could require totally different types of prompting. So mess around a bit.”
In my transient testing, V7 adhered moderately effectively to my prompts, although I didn’t have time to actually put the mannequin by the ringer.
Midjourney is an uncommon operation. Began in 2022 by Holz, who co-founded PC peripheral firm Leap Movement, it hasn’t raised a dime of outdoor cash.
In late 2023, Midjourney was reportedly anticipating to herald round $200 million in income. Lately, the San Francisco-based agency mentioned it was establishing a {hardware} workforce to work on some initiatives that it didn’t element, and it continues to coach beforehand introduced fashions for video and 3D object era.
The corporate is going through a number of lawsuits that accuse it of infringing on the rights of tens of millions of artists by coaching AI instruments on pictures scraped from the online with out the consent of the pictures’ creators.
