Overview

• V7 is Midjourney's first new AI image generation model in nearly a year, now available in alpha testing.
• The model offers improved text prompt adherence, higher image quality, and better coherence in elements like textures and hands.
• A new Draft Mode enables faster and cheaper image generation for iterative workflows, though at lower quality.
• Personalization is enabled by default, requiring users to rate 200 images to tailor outputs to their preferences.
• While V7 lacks features like upscaling and retexturing, updates are planned every one to two weeks over the next two months to address these gaps.