anima-to-z-image — anima-to-z-image v1.0

Use this two-stage image workflow to create base image content with Anima, refine it with z-image turbo, and optionally upscale the resulting image; provide either an existing image or a new latent image plus multiline text, and receive preview and saved-image outputs.

Anima creates base image content first; z-image turbo then fixes and refines the final image.

The workflow also includes steps for image sharpening, face detailing, face swapping, and upscaling.

Start with LoadImage for an existing image or EmptyLatentImage for a new latent image.

Enter multiline text with PrimitiveStringMultiline and CLIPTextEncode.

Preview results with PreviewImage or save them with SaveImage.

This is the anima-to-z-image v1.0 workflow, listed as Workflows, based on Anima, and published by Tedbiv.

Node packs listed for this workflow: comfyui_layerstyle, comfyui_ultimatesdupscale, comfyui-impact-pack, comfyui-impact-subpack, comfyui-reactor, comfyui-technodes, and rgthree-comfy.

Model files listed include 4x-UltraSharp.pth, ae.safetensors, anima-preview2.safetensors, bbox/Eyeful_v2-Individual.pt, GFPGANv1.4.pth, and GPEN-BFR-2048.onnx.

They also include qwen_3_06b_base.safetensors, qwen_image_vae.safetensors, reswapper_256.onnx, zImage_textEncoder.safetensors, and zImage_turbo.safetensors.

You can bypass the upscaler when you do not want that step.

Optional image-adjustment nodes include ImageSharpen, FaceDetailerPipe, ReActorFaceSwapOpt, and UltimateSDUpscale.

Before use, verify how to obtain zImage_textEncoder.safetensors and zImage_turbo.safetensors; their listed source URLs are blank.

Suggestion · not verified

Before use, verify how to obtain GFPGANv1.4.pth and GPEN-BFR-2048.onnx; their listed source URLs are blank.

Suggestion · not verified