Midjourney, one of the developing names in AI-generated imagery, has just launched a major upgrade, its first in nearly a year. Say hello to new Ai model (V7), now in alpha testing and already turning heads with its improved realism, smarter prompt handling, and default personalization.
This release follows closely on the heels of OpenAI’s new image generator for ChatGPT, which gained a lot of buzz for its Ghibli-inspired artwork. While Midjourney V7 isn’t officially built for Ghibli aesthetics, its output still impresses, with stunning textures, better anatomy, and richer detail across the board.
Before you can start creating with V7, you’ll need to go through a one-time onboarding process where you rate around 200 images. This isn’t just busywork, it builds your personalization profile. This helps the AI learn your unique artistic taste. This is the first Midjourney model to have personalization allowed by default. It is a step toward more tailored and creative results right out of the gate.
You can toggle V7 on through Midjourney’s web app or their Discord bot. Just look for the version drop-down and select V7. The model comes in two modes:
- Turbo Mode:- Fast and more expensive, ideal for quick prototyping
- Relax Mode:- Slower, cheaper, and great for casual exploration
There’s also a new Draft Mode, which supplies images at 10x speed and half the cost of regular generations. While the draft images are a little lower in quality, they can be improved or re-rendered with a single click.
According to Midjourney founder and CEO David Holz, new Ai model V7 is built on a completely new architecture, not just a tweak of previous versions. Holz says it delivers far better consistency when rendering human figures, hands, and complex objects, and it’s much more intelligent when interpreting both text and image prompts. However, some features like upscaling and retexturing aren’t supported just yet, but should arrive in the coming months.
“This is an entirely new model with unique strengths and probably a few weaknesses.” Holz wrote on Discord. “It may require different styles of prompting. So play around a bit.”
Related links you may find interesting
What Makes V7 Stand Out?
Compared to previous versions, V7 appears to be Midjourney’s most advanced and creative model yet. In early tests, it delivers artwork with noticeably improved coherence and finer detail, less of the janky hands and surreal smudges that often plagued earlier versions.
What’s more, this update shows Midjourney’s commitment to keeping things fresh. Unlike other AI tools that depend heavily on venture capital, Midjourney is fully bootstrapped. They’ve grown organically since launching in 2022, without taking external funding. In 2023 alone, they reportedly hit $200 million in revenue, which is remarkable for a team with a relatively small public footprint.
Beyond Images: What’s Next?
Midjourney is also focusing the next border in generative media, including video and 3D object generation. They’re even building a hardware team, though details are still under wraps. It’s clear the company isn’t just trying to keep up, it’s actively carving out a path forward in creative AI tools.
Like many companies in the space, Midjourney faces legal challenges. It’s currently the target of multiple lawsuits alleging the use of copyrighted artwork without permission during training. The outcome of these cases could set important precedents for AI training data in the creative world.
