Stable Diffusion Anime Portrait — SDXL Optimized
Generate high-quality anime portraits with optimized positive/negative prompts for Stable Diffusion SDXL and A1111.
**Positive Prompt (customize the brackets):** (masterpiece, best quality, ultra-detailed:1.3), 1girl, [AGE] years old, [HAIR COLOR] [HAIR STYLE] hair, [EYE COLOR] eyes, [EXPRESSION], wearing [OUTFIT], [BACKGROUND/SETTING], anime style, beautiful detailed face, soft lighting, cinematic composition, (8k wallpaper:1.2) **Negative Prompt (use as-is):** (worst quality, low quality:1.4), deformed, ugly, blurry, bad anatomy, extra limbs, watermark, text, (bad hands:1.3), (mutated hands:1.2), jpeg artifacts, out of frame **Recommended Settings (A1111):** - Model: Anything V5 / CounterfeitXL / Dreamshaper XL - Sampler: DPM++ 2M Karras - Steps: 28-35 - CFG Scale: 7 - Resolution: 768×1024 (portrait) - Clip skip: 2 **Useful LoRAs:** - <lora:add_detail:0.8> — sharper facial features - <lora:epiNoiseoffset:0.6> — better lighting contrast **Quick examples:** - School setting: "17yo, brown bob hair, amber eyes, school uniform, cherry blossom garden" - Fantasy: "elf princess, white long hair, violet eyes, golden crown, forest throne" - Cyberpunk: "punk girl, neon pink hair, cyber implants, rain-soaked neon city"
Result Text
Example output using school setting parameters: Positive: (masterpiece, best quality:1.3), 1girl, 17 years old, brown bob hair, warm amber eyes, gentle smile, wearing sailor school uniform with red ribbon, autumn park with golden leaves falling, anime style, beautiful detailed face, golden hour lighting, 8k wallpaper Result quality notes: - CFG 7 produces natural skin tones without oversaturation - DPM++ 2M Karras at 30 steps gives best detail-to-speed ratio - Add <lora:add_detail:0.8> for sharper eye reflections - Seed recorded: 2847563910 for reproduction The autumn lighting adds warmth to the uniform's navy blue, creating a nostalgic school memory aesthetic popular in slice-of-life anime.
评论
to leave a comment.
暂无评论，成为第一个评论的人！