Masked (Single) Workflows

This version uses SAM3 to process a single face separately for higher quality and better likeness. A detailer section masks out eyes and reprocesses them at higher quality but hands or anything else able to be selected with SAM3 can be used for the detailer as well.

v4 Workflow for Adonis Post - An additional model to be used with the Adonis Base or Refine model.

The new model was trained on the outputs of Adonis (as the control images) and the original reference high resolution images (as the target images).

This new model removes the artifacts that show up from the Base or Refine model upscale process, reducing the harsh details so the images look more natural and clean. Images of women are much softer and feminine, and the Post model seems to help with images upscaled via other means as well (SeedVR2, Upscale using Model, etc.)

The example images used Adonis_Base as the first generation step, Adonis_Refine can be used as well but will shift the image slightly (anatomy is better with Refine so what works best will vary by your image.)

There are three different workflows included: safetensors, gguf and int8.

For 30-series GPUs, the INT8 version of Flux 2 Klein 9b drastically speeds up generation times. On a 3090 using this workflow, generation times decreased from 8-9s/it to around 4-6s/it.

Download the ComfyUI-INT8-Fast node here:

https://github.com/BobJohnson24/ComfyUI-INT8-Fast

Download the flux-2-klein-base-9b-int8.safetensors here:

https://huggingface.co/vistralis/FLUX.2-klein-base-9b-INT8-transformer/blob/main/flux-2-klein-base-9b-int8.safetensors

Adonis for Flux 2 Klein 9B (Workflow)

Adonis is an "upscale model" LoKr trained using a high-resolution "target" dataset of men, paired with synthetic low-resolution edited copies as the "control." It refines skin, hair, and anatomy details that base model often gets wrong.

How it Works

• Edit-Only: Improves only what is already visible in the input image. Suitable for any image involving people.

• Two-Model Generation: The model splits into two models (adonis_base and adonis_refine) that work best together:

  1. Adonis Base: Sets the image structure and color first. (first 4-6 steps)

  2. Adonis Refine: Brings out details and corrects issues from the initial steps. (final steps, 9 steps total)

Recommended Prompt

This prompt enhances most images effectively:

uhdmanscale, fully reconstruct this entire image from cellphone quality to professional high resolution color raw quality. Remove halftone dot pattern. Apply descreen filter. Eliminate periodic grid noise. Eliminate repeating noise patterns and artifacts, remove uniform diagonal line texture patterns. Reconstruct low resolution high ISO noise areas with high resolution low ISO noise textures.

Apply full detail reconstruction to all areas: background, environment, surfaces, objects, clothing, and foreground elements — render everything sharp, textured, and high fidelity.

Subject identity is locked: preserve exact facial geometry and body geometry, eye shape and color, nose and mouth shape, and expression. On skin areas, remove color blotch artifacts, normalize tone uniformity, preserve natural pore and texture detail. On hair and body hair areas, separate smeared color artifacts, restore strand separation and texture. Outside the subject's face, freely reconstruct all texture and sharpness with no restrictions.

Deblur and focus correction pass. Infer and reconstruct underlying detail from soft source: sharpen edge definition, recover eye detail, lip definition, and skin texture from motion blur. Output as professional high resolution color camera RAW image.

Note: For non-color inputs, specify object colors in the background. Replace "color" with "black and white" if you do not want to colorize the image.

ComfyUI Setup

Install these custom nodes:

• ComfyUI-KJNodes

• ComfyUi-Scale-Image-to-Total-Pixels-Advanced

• gguf

• RES4LYF

• rgthree-comfy

• comfyui-vrgamedevgirl

Model Files & Requirements

Base Models:

• FLUX.2-klein-9B.safetensors (HuggingFace: black-forest-labs/FLUX.2-klein-9B)

• Flux 2 VAE (HuggingFace: Comfy-Org/flux2-dev)

Clip:

• Qwen3-8B-GGUF

  • Use a higher quantization GGUF if VRAM is limited. FP8 tends to degrade details in eyes and fingers based on testing.