™️ Midjourney V7 Is Here: More Speed, Quality, Details, and Personalization

Midjourney launches an alpha-version of its new V7 text-to-image AI. The model understands prompts much better, creates higher quality images (especially of people), and keeps details consistent.

Key Features:

🟠 Draft Mode: It costs half the basic price and is 10x faster, so it automatically turns on a voice mode for prompting. The quality is lower than standard, but the speed lets you quickly test ideas.

🟠 Turbo and Relax Modes: Turbo is faster but costs twice as much as the standard v6 generation. Relax is half the price but slower. A standard mode will be added soon.

🟠 Personalization is on by default in V7. To start using the model, you'll need to rate 200 images to set up your profile (takes about 5 minutes). You can toggle off this feature later.

📎 Midjourney highlights that V7 is a completely new model. Although it has its strengths, it may also have weaknesses and might require a different approach to prompting.

V7 still supports the existing --sref styles, but the way they are understood may have changed slightly.

This isn't the final release yet. During the next two months, the developers will introduce new features every week or two. They also promise better ways to work with reference images.

Meanwhile, users are already sharing the first results from the new model ⤴️

What do you think about the Midjourney V7?

❤️ — awesome, much better!
🤔 — not bad, but not much progress.
🎃 — it actually got worse...

#news #Midjourney @hiaimediaen