Midjourney V6.1 is here, and this update truly stands out. If you’ve been using V6.0, you’ll notice major gains in image coherence and detail handling, especially with human limbs and facial features. Below is a breakdown of the key new features to help you get up to speed quickly.
Major Coherence Boost – Say Goodbye to Distorted Limbs
Distorted arms, legs, and other body parts have long been a common issue in AI-generated art. V6.1 specifically addresses this pain point, producing more natural body structures. Even when characters appear small in the frame, facial details remain crisp — the old “blurry small face” problem is essentially gone. Overall visuals look notably better, with more vibrant colors.
More Accurate Text Generation & Better Prompt Alignment
Many Midjourney users want to generate images with embedded text, but earlier versions struggled with accuracy. V6.1 improves English text generation — when you use quotes to specify the displayed text, the results are far more reliable than in V6.0. Additionally, the output aligns more closely with your prompts, reducing the chance of mismatched or irrelevant imagery.
New Parameter Mode – Greater Flexibility Between Speed & Detail
V6.1 reintroduces the --q 2 parameter. In standard mode, image generation is about 25% faster, significantly cutting wait times. If you need richer texture detail, try using --q 2 — it uses roughly 25% more GPU time for finer quality, though it may slightly affect image coherence. Conversely, --q 0.5 is ideal for quickly testing prompts during the concept phase.
