Azure Luminara
Azure Luminara serves as a universal base model under the NoobAI framework
Model Merge Composition
🐾 YiffyMix - v62-NoobXL | NoobAI Checkpoint | Civitai
🦊 Indigo Furry Mix XL - NoobAI_EPS_Ⅲ_3 | NoobAI Checkpoint | Civitai
❄️ Azure ColdSnap - ColdSnap | Illustrious Checkpoint | Civitai
This hybrid combination creates a well-balanced foundation, supporting diverse art styles from stylized anthro and fantasy to anime and semi-realistic rendering.
Dataset Information
The model incorporates a curated dataset sourced from e621,
Balance anatomy between furry and humanoid subjects
Improve cross-style adaptability (e.g., anthro–human hybrids)
Performance Notes
Works seamlessly with both NoobAI and Illustrious model structures
Balanced rendering across furry and non-furry characters
Maintains strong color grading, anatomical accuracy, and background composition
🏷️ Prompt Tags
✅ Positive (Quality Boost):
masterpiece, best quality, ultra-detailed, 8k, soft lighting, illustration, anime style, clean background, highly detailed eyes
🚫 Negative (Suggested):
lowres, (worst quality:1.2), (bad quality:1.2), (low quality:1.2), bad anatomy, deformed hands, overexposure
Description
Luminara Core Update Changelog
This update focuses on making Luminara Core more reliable and stable
Base Model
Base model merge: unchanged
The core structure and overall look of Luminara remain the same.
What’s New
Added a LoRA merge specifically to:
Stabilize under-exposed / dark images
Improve readability in night scenes
Reduce small anatomy issues in low-light setups
Visual Impact
Better exposure control in dark or moody lighting
Characters feel more stable anatomically, with cleaner silhouettes
Less visual “muddiness” in shadows and low-contrast areas
Cleaner backgrounds with less visual noise and distraction