If you're a heavy Midjourney user, you've likely noticed several eye-catching new features rolling out over the past few weeks. From smarter personalization settings to a fully functional image editor, these updates make AI image generation more controllable and enjoyable. Let's dive into the most notable changes in Midjourney's latest features so you can hit the ground running.
Personalization Models and Style Customization
Among the new Midjourney features, improvements to personalization are a top priority for many users. Now, simply rate images on the official website—after at least 200 ratings—to unlock your own preference model. Add the --p parameter to your prompts, and the results will better match your aesthetic taste. This model continuously learns from your rating activity. Around the same time, Style Reference has been upgraded to V2, offering finer visual style anchoring. Combined with the Multiple CREFs feature, you can preserve facial features across multiple characters simultaneously.
Image Editor Redefines the Workflow
The web-based image editor is another game-changing update. Previously, local retouching or preserving specific areas required external editing tools. Now, Midjourney includes a full editing panel where you can use a brush to select areas for modification or mark regions to keep unchanged. This feature performs exceptionally well on the V6.1 model. Paired with the Expand tool, it intelligently fills in edge content. Note that when editing faces, using the character reference parameter helps maintain stability.
