Z Image Turbo Workflow by Stable_Yogi🔥 — ZimageTurbo+XL V1

This ComfyUI workflow is for generating and refining Z Image Turbo images from text prompts, a seed, and latent controls; it can apply LoRA and ControlNet guidance and produces preview or saved images.

Processing stages include text encoding, model sampling, VAE decoding, ControlNet, face detailing, image comparison, and upscaling.

The supplied model files are listed as a text encoder, UNet, and VAE: qwen_3_4b.safetensors, z_image_turbo_bf16.safetensors, and ae.safetensors.

You can enter text, set a seed, create a latent image, choose LoRA loaders, apply ControlNet, and pass images into detailing or upscaling stages.

Outputs include PreviewImage and SaveImage, with Image Comparer (rgthree), FaceDetailer, and UltimateSDUpscale stages for reviewing or processing images.

Before use, make the named ComfyUI model files and node packs available.

Sugestão · não verificado

The workflow names these node packs: comfyui_tinyterranodes (2.0.9), comfyui_ultimatesdupscale (1.6.0), comfyui-easy-use (1.3.4), comfyui-impact-pack (c6056b132d7e155c3ece42b77e08ea45bde1bfef), comfyui-impact-subpack (74db20c95eca152a6d686c914edc0ef4e4762cb8), efficiency-nodescom

Named model files include z_image_turbo_bf16.safetensors, qwen_3_4b.safetensors, ae.safetensors, qinglong_detailedeye_z-imageV2(comfy).safetensors, Realism Lora By Stable Yogi_V3_Lite.safetensors, and 4xUltrasharp_4xUltrasharpV10.pt.

Other named files include sam_vit_b_01ec64.pth, Xinsir_Controlnet_Union_for_ SDXL_promax.safetensors, bbox/face_yolov8n.pt, bbox/Eyes.pt, Realism Lora By Stable Yogi_V3_Pro.safetensors, and Realism BSY Pony V6 Patreon Pro DMD2 C42.safetensors.

The supplied ComfyUI paths are models/text_encoders/qwen_3_4b.safetensors, models/loras/qinglong_detailedeye_z-imageV2(comfy).safetensors, models/diffusion_models/z_image_turbo_bf16.safetensors, and models/vae/ae.safetensors.

The same notes also list z_image_turbo-Q4_K_S.gguf under models/diffusion_models.

Before use, verify the exact source, filename, and file type for sam_vit_b_01ec64.pth, Xinsir_Controlnet_Union_for_ SDXL_promax.safetensors, bbox/face_yolov8n.pt, bbox/Eyes.pt, Realism Lora By Stable Yogi_V3_Pro.safetensors, and Realism BSY Pony V6 Patreon Pro DMD2 C42.safetensor

Sugestão · não verificado

The release is labeled ZimageTurbo+XL V1, and the notes point to the workflow notes and a Discord community for help.