FLUX - text2img + upscale workflow for beginner (2 option included)

Use this ComfyUI workflow to turn a text prompt into a FLUX-generated image and then an upscaled image; use the same prompt for both steps or enter a separate upscaling prompt.

The workflow uses sequential phases, with image generation followed by upscaling.

Input a generation text prompt and, if desired, a different prompt for upscaling. The outputs are a generated image and an upscaled image.

Use it in ComfyUI; Flux.1 D is listed as the base model.

Listed model files include 1xSkinContrast-HighAlternative-SuperUltraCompact.pth, clip_l.safetensors, FLUX_VAE_diffusion_pytorch_model.safetensors, flux1-dev-Q8_0.gguf, flux1DevFp8_v10.safetensors, remacri_original.pth, and t5xxl_fp16.safetensors.

Listed node types include CLIPTextEncodeFlux, DualCLIPLoader, EmptyLatentImage, Image Comparer (rgthree), KSampler, PreviewImage, and SaveImage.

Other listed node types include UltimateSDUpscale, UNETLoader, UnetLoaderGGUF, UpscaleModelLoader, VAEDecode, and VAELoader.

For upscaling, choose the generation prompt again or enter a different prompt.

Before using it, verify that the listed model files and node types are available in ComfyUI.

Vorschlag · nicht geprüft

A license is not listed, so check the Civitai page for usage terms before reuse.

Vorschlag · nicht geprüft