HS-Anima 4K Final Output Workflow — LoRA+Face Fix+Upscale

This Anima workflow is for creating anime images with multiple LoRAs, repairing faces and enhancing detail, and upscaling to HD/4K; it names text encoding and latent-image setup among its stages, plus image preview and saving.

Named stages include CLIPLoader, CLIPTextEncode, EmptyLatentImage, KSampler, FaceDetailer, LatentUpscaleBy, VAEDecode, PreviewImage, and SaveImage.

The listed base model is Anima, and the version is LoRA+Face Fix+Upscale.

PreviewImage and SaveImage are the named stages for previewing and saving images.

The Civitai listing is titled HS-Anima 4K Final Output Workflow — LoRA+Face Fix+Upscale, lists Hshi as publisher, and is at https://civitai.com/models/2679939?modelVersionId=3009170.

Listed model files are 4x-UltraSharp.pth, bbox/face_yolov8m.pt, hs-anima-1.0.safetensors, LineLore1anima.safetensors, qwen_3_06b_base.safetensors, qwen_image_vae.safetensors, and sam_vit_b_01ec64.pth.

Other named components include CLIPLoader, CLIPTextEncode, EmptyLatentImage, FaceDetailer, Fast Groups Muter (rgthree), KSampler, LatentUpscaleBy, LoraLoader, MaskToImage, PreviewImage, PrimitiveStringMultiline, SAMLoader, SaveImage, UltimateSDUpscale, UltralyticsDetectorProvider

Other named components include UNETLoader, UpscaleModelLoader, VAEDecode, and VAELoader. Listed requirements also include qwen_image_vae.safetensors, qwen_3_06b_base.safetensors, bbox/face_yolov8m.pt, ComfyUI-Impact-Pack, and ComfyUI-Impact-Subpack.

The workflow is presented with Multi-LoRA support, Flux2Klein enhancement, and SeedVR2 4K upscaling.

It is also presented for anime-character face repair, face-detail enhancement, and HD upscaling.

Before downloading, confirm the title, publisher, and version match what you intend to use; the listing does not specify a license, so check usage terms.

Suggestion · non vérifié

Check how to obtain LineLore1anima.safetensors and hs-anima-1.0.safetensors before setup; no provider is named for either file.

Suggestion · non vérifié