ComfyPilotComfyPilot
模型库 / LoRA / [MOSS|H3 image] onineko (おにねこ) style
[MOSS|H3 image] onineko (おにねこ) style

[MOSS|H3 image] onineko (おにねこ) style

3008630

LoRA 447 次下载 30 喜欢 更新于 2026-11-14

底模：MiniMax H3

style

下载模型 在源站打开

触发词

onineko

文件清单 线路：镜像（hf-mirror.com） · 切换

下载后放进对应的 ComfyUI 目录。models/ 指 ComfyUI 的模型根目录。

权重 onineko_h3_v2.safetensors 296 MB 下载

说明

[MOSS|H3 image] onineko (おにねこ) style

作者：KaleidRuby

onineko style for MOSS 0.6-1.0, cfg=3, steps=50 Trained on base, works also for SFT (higher intensity needed) For minimax H3 IMAGE, NOT H3 video recommand setting 16MP, cfg=1, step=6 recommand strength: 0.8-1.2 Note: H3 image model will reproduce bad image when the total lora strength (including other loras) is above 2.0, improve steps from 6 to 12 will help solve the problem

示例图