Home Creators Posts Import Register Favorites Logout

Videos

  • 589774060.mp4

Downloads

Content

Recent video tutorial for Qwen Image 2512 : https://www.youtube.com/watch?v=frv9YGRtWRQ

This workflow are using Qwen Image 2512 a little more advanced and generate image up to 4K with enhance sampling.

The video on the top are recorded raw footage of the process of making this workflow, and steps of using it. Pay attention of how to use it, if you are not sure.

Something need to pay attention on....

Create text prompt

can be use Qwen3VL or Bypass Qwen3VL use your own prompt only.

Qwen3VL reference image can be use by image URL or file path.

If Qwen3VL Bypass, you can use the textbox on the right size to type your own text prompt. That means the image generation are using your input text prompt ONLY for style reference.

Then it will be generate.

Currently I have not enable any LoRAs in the workflow by default, you can enable it on your own choice.

Model Links :

Qwen-Image-2512-BF16.safetensors (merge by myself)

https://huggingface.co/benjiaiplayground/Qwen-Image-2512-BF16/tree/main

Text encoder and VAE remain the same as previous Qwen Image AI Model

Comfy-Org - qwen_image_2512_bf16.safetensors

https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/main/split_files/diffusion_models/qwen_image_2512_bf16.safetensors

Comfy-Org - qwen_image_2512_fp8_e4m3fn.safetensors

https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/main/split_files/diffusion_models/qwen_image_2512_fp8_e4m3fn.safetensors

text_encoders

https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main/split_files/text_encoders

vae

https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main/split_files/vae

GGUF : https://huggingface.co/unsloth/Qwen-Image-2512-GGUF

Qwen-Image-2512-Lightning https://huggingface.co/lightx2v/Qwen-Image-2512-Lightning/tree/main

*** update: sampler setting , feel free to adjust according to your image style. SeedVR2 upscaler, use models that nott label as sharp preform smoother for image gen.

Have Fun! :)

Files

Comments

Tethyan

Thanks Benji, Excellent Workflow ,, please add model unload checks as its consume more VRAM.