<h1 id="phaenon-character-lora-(animagine-xl-sdxl)">Phaenon — Character LoRA (Animagine XL / SDXL)</h1><p>High-quality character LoRA for Phaenon, trained on a clean high-resolution dataset. Designed specifically for Animagine XL 3.1 and compatible SDXL anime checkpoints.</p><p>It captures his facial features, light blue eyes, signature white/silver hair with subtle blue gradients, and detailed outfit elements with high consistency, while offering excellent flexibility for clothing customization.</p><h2 id="recommended-settings">Recommended Settings</h2><p><strong>ParameterRecommended Value</strong><span style="font-family:&quot;Google Sans Text&quot;, sans-serif;line-height:1.15">Base ModelAnimagine XL 3.1 / Any SDXL Anime CheckpointLoRA Weight0.8 (optimal range: 0.7 - 0.9)Trigger WordphaenonHelper Tags1boy, solo, blue eyes, white hair, light blue hair tipsCFG Scale5.0 - 7.0Sampling Steps25 - 30SamplerEuler a or DPM++ 2M KarrasSchedule TypeKarrasResolution832x1216 (Portrait) / 1216x832 (Landscape)</span></p><h2 id="key-features">Key Features</h2><ul><li><p>Outfit Flexibility: Clean tag separation allows swapping his default gear for casual, modern, or alternative clothing without style bleeding.</p></li><li><p>Pose Stability: Performs reliably across portraits, full-body shots, and dynamic action angles.</p></li><li><p>High Facial Consistency: Maintains accurate eye color, hair shape, and facial features across different weights.</p></li></ul><h2 id="example-prompts">Example Prompts</h2><h3 id="positive-prompt-template:">Positive Prompt Template:</h3><p><span style="font-family:&quot;Google Sans Text&quot;, sans-serif;line-height:1.15">Plaintext</span></p><pre><code>score_9, score_8_up, score_7_up, &lt;lora:Phaenon:0.8&gt;, 1boy, phaenon, solo, looking at viewer, upper body, casual clothes, outdoor, masterpiece, best quality, highly detailed
</code></pre><h3 id="negative-prompt-template:">Negative Prompt Template:</h3><p><span style="font-family:&quot;Google Sans Text&quot;, sans-serif;line-height:1.15">Plaintext</span></p><pre><code>score_4, score_3, score_2, score_1, lowres, bad anatomy, bad hands, missing fingers, extra digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, blurry
</code></pre><p>Feedback, reviews, and generations shared in the gallery are always appreciated.</p>

Files with SHA256: 6fb1b7a7216239aabb35463c100268ad8b1903a913f424d952493351bf645c9d