Gladas' Workflow for Krea 2 Turbo — v1 — LC Speed v2

Use this ComfyUI workflow for Krea 2 text-to-image generation with the LC Speed setup; it supports multiple LoRA loading, wildcards and upscaling, using text prompts to produce images.

Sampler settings are shared with Image Saver, and ModelSamplingAuraFlow is included.

Saved images can include full metadata in a format readable by CivitAI.

If you use an AIO model, add a checkpoint loader in place of the removed AIO Loaders and reconnect the nodes.

Sugestão · não verificado

The workflow lists these node packs: bkwildcards, ComfyMath, ComfyUI_LC123_nodes, comfyui-custom-scripts, comfyui-image-saver, comfyui-impact-pack, comfyui-impact-subpack, comfyui-kjnodes and rgthree-comfy.

Its listed model files include bbox/Eyeful_v2-Individual.pt, betterDaysIllustriousXL_illuXLV02BAKEDVAE.safetensors, qwen_image_vae.safetensors and qwen3vl_4b_fp8_scaled.safetensors.

Other listed model files are sam_vit_b_01ec64.pth, seedvr2_3b_int8_convrot.safetensors, seedvr2_ema_vae_fp16.safetensors and vxpKrea2Nsfw_beta3INT8.safetensors.

The LC Speed setup includes an option for its High Pass/Low Pass version.

The workflow is not set up for models with CFG above 1, and detailers are listed as removed on v12.

Before setup, check which ComfyMath package the workflow expects; Anima Turbo settings are left for you to determine.

Sugestão · não verificado