<h1 id="glossy-mature-anime-style">Glossy Mature Anime Style</h1><p>A distilled style LoRA for <strong>mature, high-gloss dark-fantasy anime aesthetics</strong> — think dramatic specular highlights on skin and materials, moody cinematic lighting, and confident, adult character design.</p><p><strong>Trigger word:</strong> <code>thedaves_glossy_mature</code></p><hr /><h2 id="what-it-does">What it does</h2><ul><li><p>Strong glossy / wet-look rendering on skin and materials — no extra prompt words needed, it's baked into the trigger</p></li><li><p>Mature facial styling: defined features, confident expressions</p></li></ul><h2></h2><p>for weight, i use 1.0, seems to work :D</p><p></p><h2 id="example-prompt-(anima)">Example prompt (Anima)</h2><pre><code>thedaves_glossy_mature

masterpiece, best quality, newest, sensitive,
a curvy tan-skinned woman with medium-length wavy auburn hair and hazel eyes, standing on a quiet balcony at dusk, warm orange sky, potted plants around her, relaxed confident smile, wearing a simple fitted top and jeans,
upper body, from front,</code></pre><p>Swap in your own character LoRas, Settings, poses (SFW / NSFW both work). I would love to see your creations and improve this style further!</p><h2 id="notes">Notes</h2><p></p><ul><li><p>v1 release — feedback and results welcome.</p></li><li><p>The Krea 2 Version works well for NSFW if you use the Mystic XXX LoRa. It creates shiny, semi realistic pictures of wet skin. Nice byproduct, if you ask me :)</p></li></ul><p></p>

Glossy Mature Anime Style

<p>A Krea 2 (RAW-trained, Turbo-compatible) version of the glossy, high-contrast mature anime style — same aesthetic direction as the original Anima release, trained natively for Krea 2's architecture.</p><p><strong>Trigger word:</strong> <code>thedaves_glossy_mature</code></p><p>Settings</p><ul><li><p>Weight: 1.0</p></li><li><p>Turbo checkpoint, 8 steps, CFG 0–1, sampler <code>er_sde</code> / scheduler <code>simple</code></p></li><li><p>Prompt in natural-language prose — no tags, no <code>(word:1.2)</code> weighting (Krea 2's encoder doesn't read it)</p></li></ul><p>Important: NSFW</p><p>Krea 2's base model has safety filtering built into training. This LoRA controls <strong>style</strong>, not content filtering — for NSFW results you'll also need a checkpoint variant without the built-in filter, or a conditioning-rebalance / "filter" LoRA (layer-reweighting setups like ConditioningKrea2Rebalance). Without one of those, explicit prompts may get filtered regardless of this LoRA.</p><p>v1 release — feedback welcome.</p>

Files with SHA256: 855b8ba4f7c39cf4a869e169683f6783d049803da596942a72792d6e81290fa6