The holiday season has begun, and Midjourney is giving away presents.
Midjourney which is an "independent research lab exploring new mediums of thought and expanding the imaginative powers of the human species," is a versatile text-to-image service, popular for generating AI artwork in various scenarios like logos, portraits, and architecture.
The popular image generation AI model from the research collective of the same name founded by David Holz, has just announced Midjourney version 6.
Like a gift out of no where, version 6 of this powerful AI packs massive improvements.
For starters, among those new features are drastically improved and more realistic, highly detailed images, and the ability to have the model generate legible text within images.
This particular feature is something that is quite difficult to achieve in generative AI products, including Midjourney.
Earlier versions of Midjourney, like many AI image tools, struggled with creating clear and coherent text, often resulting in garbled or nonsensical characters. The introduction of this feature in version 6 is a significant leap, opening up new possibilities for designers and artists who wish to incorporate textual elements into their AI-generated visuals.
Read: How Midjourney AI Art Generator Creates Photorealistic Images, Without The Nightmarish Hands
We're now alpha-testing our V6 models Midjourney. Just type /settings and click V6 or add --v 6 after your prompt. Image coherence and prompt understanding are greatly improved. You can draw text and dolphins and there's new upscalers too. Happy holidays everyone!
— Midjourney (@midjourney) December 21, 2023
Then, one of the most notable feature, is an entirely new prompting method has been developed.
Midjourney’s prompting allows users to generate images by typing in specific text descriptions and keywords into the Discord server or alpha version of the website.
Traditionally, the process has been esoteric and technical, with users sharing examples of techniques that had worked well for them on social media.
Users can also place all kinds of elements in the generations, like frames, and even text.
Initially introduced in an alpha release, version 6 also has a more accurate prompt following as well as longer prompts, improved coherence, and model knowledge, and improved image prompting and remix. Prompts can also be much longer, up to more than 350 words, and that users can also specify colors and other details.
It also improved its upscalers, with both 'subtle‘ and 'creative‘ modes.
With version 6, Holz made it clear on his Discord post, stating that prompting tricks would no longer result in the type of results users desired.
"You will need to re-learn how to prompt," he wrote.
"Prompting with V6 is significantly different than V5. You will need to ‘relearn’ how to prompt."
In this case, if users want something more photographic or less opinionated or more literal, they you should probably default to using --style and the --stylize.
This happens because version 6 is much more sensitive to prompts.
What this means, users are advised to avoid generic "award winning, photorealistic, 4k, 8k" prompts.
Users have to be more explicit about what you want.
Long story short, Midjourney version 6 is capable of adapting much better to users descriptions (or prompts).
When writing what we want to generate, we can “talk” to Midjourney in the style of how we would with ChatGPT. It is capable of better understanding natural language.
"This model can generate much more realistic imagery than anything we’ve released before," wrote Holz in a message posted in the Midjourney Discord server, which has over 17 million members.
Holz also said that Midjourney version 6 was actually the "third model trained from scratch on our AI superclusters," and took nine months to develop.
At this time, Holz noted in his Discord message announcing V6, that the model "is an alpha test," meaning that things can change frequently and without prior notice.
" [...] V6 isn’t the final step, but we hope you all feel the progression of something profound that deeply intertwines with the powers of our collective imaginations," he said,
The announcement calls for users to employ these "incredible powers with joy, wonder, responsibility, and respect," which has always been part of Midjourney’s ethos.
"Don't be a jerk or create images to cause drama," the announcement reads.