Remove Watermarks from images in ComfyUI (Workflow Included)
Watermarks ruin an otherwise good image. This ComfyUI workflow uses the Flux model to repaint the image at a low denoising strength, removing the watermark while keeping the rest of the picture largely unchanged. Original with watermark. Watermark removed! In this tutorial, you will learn how to…
Annons
Annons
Watermarks ruin an otherwise good image. This ComfyUI workflow uses the Flux model to repaint the image at a low denoising strength, removing the watermark while keeping the rest of the picture largely unchanged. Original with watermark. Watermark removed! In this tutorial, you will learn how to Set up the Flux inpainting workflow in ComfyUI. Remove a watermark from an image. Tune the denoising strength to balance cleanup and fidelity. Disclaimer: Only remove watermarks from images you own or have permission to edit. Software needed We will use ComfyUI, a free AI image and video generator. You can use it on Windows, Mac, or Google Colab. Think Diffusion provides an online ComfyUI service if you don’t have a capable local GPU. They offer an extra 20% credit to our readers. How the workflow works The workflow is a standard Flux inpainting setup with a twist: instead of painting a mask by hand, it builds the mask from the colors of the input image. The input image is resized to 1024 pixels. Create a white mask (i.e. inpaint the whole area) for the image. Inpaint the whole image using the Flux.1 Fill model. Critical note: Don’t use the standard Flux.1 Fill model. It doesn’t work well for whole-image inpainting. Use the finetuned OneReward variant. Because the prompt is empty, Flux fills the area based on the surrounding image content alone. Watermark remover workflow Step 0: Update ComfyUI Before loading the workflow, make sure your ComfyUI is up to date. The easiest way is to use ComfyUI Manager. Click the Manager button on the top toolbar. Select Update ComfyUI. Restart ComfyUI. Step 1: Download the models Download these models: Download flux.1-fill-dev-OneReward-transformer_fp8.safetensors and put it in ComfyUI > models > diffusion_models. Download clip_l.safetensors and put it in ComfyUI > models > text_encoders. Download t5xxl_fp16.safetensors and put it in ComfyUI > models > text_encoders. Download ae.safetensors (Flux VAE) and put it in ComfyUI > models > vae. After loading the workflow, select flux.1-fill-dev-OneReward-transformer_fp8.safetensors in the Load Diffusion Model node. If you already use Flux on ComfyUI, you likely have the CLIP, T5 and VAE files. Step 2: Load the workflow Download the workflow JSON file below. Drop it into ComfyUI to load. Download Step 3: Upload an image Upload the image with the watermark to the Load Input Image node. You can use the test image below. Step 4: Run the workflow Click the Run button to run the workflow. Watermark removed. Settings SettingValue in the workflowEffectResize (longest side)1024Working resolution. Larger images are downscaled.Steps20Sampling steps.Sampler / schedulereuler / normalDenoise0.4Higher removes more but changes more of the image.Flux Guidance30Prompt/inpaint adherence.Seed707577193281429 (fixed)Change it to get a new result. Tips If the watermark is still visible, increase denoise in small steps (0.4 → 0.5 → 0.6). If the image changes too much, decrease denoise. Change the seed to generate a new variation. The image is resized to 1024 pixels. Upscale the result afterward if you need the original resolution. What’s next Qwen Image Edit Inpaint How to remove a person with AI inpainting Flux Fill inpainting Inpainting in ComfyUI AI upscalers The post Remove Watermarks from images in ComfyUI (Workflow Included) first appeared on Stable Diffusion Art.