🌀 Visual Regional Prompting – Multi-Model Workflows, Native LoRA Hooks & LLLite

## Anima Native Regional LoRAThis experimental workflow demonstrates an external LoRA applied only to a selected Anima region.The included test assigns <code>Skin-tone-Slider-Anima</code> at <code>+6</code> to the left subject while the touching subject on the right remains an unmodified control. It uses BV Regional Native Conditioning in <code>hybrid</code> mode with a tested blend ratio of <code>0.35</code>.### Requirements- BV Node Pack <code>v0.10.0</code> or newer- ComfyUI-Lora-Manager for the included <code>LORA_STACK</code> node- Compatible Anima model, text encoder and VAE- <code>Skin-tone-Slider-Anima</code> from Civitai RedModel weights and LoRAs are not included.### ImportantRegional LoRA support currently applies to BV's native-conditioning path, not the model-specific Attention nodes. Regional guidance and LoRA isolation remain model-, training-, prompt- and seed-dependent.Basic compatibility was also tested with Illustrious, but those results did not meet the quality bar for a separate showcase workflow.

VisualRegionalPromptingMulti_animaLoraNativeEXP.json

Anima Lora Native (EXP)

Krea 2 Lora Attn (EXP)

Anima Lora Attn (EXP)

Krea 2 (Experimental)

Flux.2 klein 9B

ZImage Turbo

Illustrious/Pony/SDXL

Anima

Files with SHA256: c308f8231ea77de2eebefdb94b273cb2c4dd6387f0b3ac9a0dff90e358379954