Midjourney Unveils V7: A Revolutionary AI Image Generation Model

Midjourney, one of the pioneering AI-based image generators on the internet, has unveiled its first upgraded image model, developed using AI technologies, nearly a year later.

The model, known as V7, began its alpha testing phase around midnight EST on Thursday, just a week after OpenAI launched a new image generator in ChatGPT that quickly became viral due to its ability to create images in the style of Ghibli.

David Holz, CEO of Midjourney, described V7 as a «completely different architecture» in a post on X.

“V7 […] is much smarter when it comes to text prompts,” Holz elaborated in an announcement on Discord. “The prompt images look fantastic, with noticeably improved quality thanks to beautiful textures, and the alignment of bodies, hands, and various objects is significantly better in all details.”

V7 comes in two options: Turbo (which is more resource-intensive) and Relax. It features a new tool called Draft Mode, which renders images ten times faster and at half the cost compared to the standard mode. While images in Draft Mode are of lower quality than those in the standard mode, they can be enhanced and re-rendered with a simple click.

According to Holz, several standard features of Midjourney are currently unavailable for V7, including image scaling and texturing. He mentioned that these functionalities will likely be introduced soon, potentially within two months.

«This is a completely new model with unique advantages and probably a few drawbacks,» Holz stated on Discord. «We want to hear from you about what works well and what doesn’t, but please keep in mind that it may require different styles of prompts. So, experiment a bit.»

During brief testing, V7 responded fairly well to prompts, though there wasn’t enough time for a thorough evaluation of the model.

Midjourney is an unusual startup. Founded in 2022 by Holz, who co-founded the PC peripheral company Leap Motion, it has not raised a dime from external sources.

By late 2023, Midjourney reportedly anticipated generating around 200 million dollars in revenue. Recently, the San Francisco-based company stated that it is forming a hardware development team for unspecified projects and continues to train previously announced models for video and 3D object creation.

The company has faced several lawsuits alleging that it violated the rights of millions of artists by training its AI tools on images downloaded from the internet without the creators’ consent.

Source