Midjourney’s latest round of updates focuses on solidly delivering three things: “more accurate outputs, more stable details, and shorter waiting time.” This article breaks down Midjourney’s new changes by use case, and gives you settings and prompt-writing formulas you can copy directly.
More natural structural consistency: character limbs no longer depend on luck
When many people use Midjourney to generate characters, their biggest fear is fingers, arms, or legs breaking or having strange proportions. The new version is more stable in body-structure continuity, with a more unified overall character form—especially with fewer failures in “limb connections” between people and animals/plants.
In practice, you can describe the “pose” more specifically to reduce the model’s improvisation: add descriptions like “full body, arms down, hands visible” in your Midjourney prompt, and pair them with clearer shot terms (such as “medium shot / full shot”) to increase the success rate.
Detail and texture upgrades: clearer skin, materials, and small objects in the distance
Midjourney has noticeably improved texture, skin feel, and image layering, while producing fewer pixel artifacts. Common issues from before—“blurry eyes, weird hands in the distance, lost tiny facial details”—are now more likely to be right on the first try in Midjourney.
If you make product images or realistic portraits, it’s recommended to add material and lighting to your Midjourney prompts, such as “soft studio lighting, realistic skin texture, fine fabric weave,” to take advantage of the upgraded detail.
A new upscaler and faster generation: more output in the same amount of time
This time Midjourney also introduced a new upscaling direction that emphasizes image and texture quality; at the same time, standard generation tasks are faster, letting you see results sooner and iterate more quickly. For design workflows that require frequent drafts, “waiting a little less” in Midjourney can save an entire round of communication time.
