Anima img2img (denoise) with Loras — v1.0

Anima img2img (denoise) with Loras is a workflow for image-to-image denoising; it takes an input image, includes image sizing, and provides preview and saved-image outputs.

The listed stages include image loading and sizing, VAE encoding and decoding, sampling, LoRA loading, previewing, and saving.

The workflow lists Anima as its base model and requires Lenovo UltraReal for lenovo_anima.safetensors.

Provide an input image through LoadImage; ImageScaleToMaxDimension is included for sizing.

PreviewImage is included for previewing, and SaveImage is included for saving images.

Find the workflow at https://civitai.com/models/2595006; the listed publisher is Sinatra.

The named Anima files are anima-preview3-base.safetensors, qwen_3_06b_base.safetensors, and qwen_image_vae.safetensors.

Other named model files are anime-studio-1-anima-fp8.safetensors and lenovo_anima.safetensors.

Node types include CLIPLoader, CLIPTextEncode, ImageScaleToMaxDimension, KSampler, LoadImage, LoraLoaderModelOnly, MarkdownNote, PreviewImage, SaveImage, UNETLoader, VAEDecode, VAEEncode, and VAELoader.

The workflow includes LoraLoaderModelOnly for LoRA handling and ImageScaleToMaxDimension for image sizing.

Model-loading nodes include UNETLoader and VAELoader.

Before use, verify that anime-studio-1-anima-fp8.safetensors and lenovo_anima.safetensors are available.

Vorschlag · nicht geprüft

No license is listed for this workflow, so check https://civitai.com/models/2595006 before reuse.

Vorschlag · nicht geprüft