<h3 id="workflow-overview">Workflow Overview</h3>This is an efficient and powerful ComfyUI workflow designed for high-quality image outpainting using the Qwen Image Edit model. Outpainting allows you to extend an image beyond its original borders, seamlessly generating new content that matches the style, lighting, and context of the original. This workflow is optimized for speed and simplicity, delivering impressive results in just a few steps.Core Concept: Upload an image, and the AI will intelligently expand the canvas, generating plausible and coherent extensions of the scene while perfectly blending the new content with the original.<hr /><h3 id="key-features-and-highlights">Key Features &amp; Highlights</h3><ul><li>🤖 Specialized Model: Utilizes the <code>Qwen_Image_Edit-Q5_0.gguf</code> model, specifically fine-tuned for image editing tasks like outpainting and inpainting.</li><li>⚡ Lightning Fast: Integrates the <code>Qwen-Image-Edit-Lightning</code> LoRA, enabling high-quality outpainting results in only 8 sampling steps.</li><li>🎯 Precision Prompting: Uses dedicated <code>TextEncodeQwenImageEdit</code> nodes that understand the image context, ensuring the model follows instructions for seamless extension.</li><li>🖼️ Automated Preprocessing: The workflow automatically pads your image to create space for outpainting and scales it to an optimal size for the model.</li><li>🔧 Optimized Pipeline: Pre-configured with expert negative prompts and optimal settings (CFG, sampler) for outpainting, so you get great results by default.</li><li>📁 One-Click Operation: Just load your image and run the workflow. No complicated settings need to be adjusted.</li></ul><hr /><h3 id="workflow-structure">Workflow Structure</h3>The workflow is neatly grouped into logical sections for easy understanding and customization:<ol><li><code>Step1 - Load models</code>: Loads the main Qwen Image Edit model, its specialized CLIP vision encoder, and the VAE.</li><li><code>Step 2 - Upload image for editing</code>: Loads your input image and preprocesses it for outpainting (padding and scaling).</li><li><code>Step 3 - Prompt</code>: Where the AI is given instructions on how to outpaint the image, with pre-written positive and negative prompts for optimal results.</li><li>Sampling &amp; Decoding: The KSampler runs for 8 steps with Euler, and the VAE decodes the latents into the final outpainted image.</li><li>Image Output: The <code>SaveImage</code> node saves the final result.</li></ol><hr /><h3 id="how-to-use-this-workflow">How to Use This Workflow</h3><ol><li>Download &amp; Install:<ul><li>Ensure you have ComfyUI Manager to easily install missing custom nodes.</li><li>Required Custom Nodes: <code>ComfyUI-GGUF</code> (for loading the .gguf models).</li><li>Download the <code>.json</code> file from this post.</li></ul></li><li>Load the Models:<ul><li>Main Model: Place <code>Qwen_Image_Edit-Q5_0.gguf</code> in your <code>ComfyUI/models/gguf/</code> folder.</li><li>CLIP Model: Place <code>qwen2.5-vl-7b-it-q4_0.gguf</code> in the same <code>gguf/</code> folder.</li><li>VAE: The workflow points to <code>qwen_image_vae.safetensors</code>. Ensure it's in your <code>models/vae/</code> folder.</li><li>LoRA: Place <code>Qwen-Image-Edit-Lightning-8steps-V1.0-bf16.safetensors</code> in your <code>models/loras/</code> folder. Adjust the path in the LoraLoader node if yours is in a subfolder (e.g., <code>qwen_loras/</code>).</li></ul></li><li>Load Your Image:<ul><li>In the <code>LoadImage</code> node, change the image name to your own file (e.g., <code>my_landscape.jpg</code>).</li></ul></li><li>Customize the Outpainting (Optional):<ul><li>The positive prompt is pre-configured for general outpainting. For specific requests (e.g., "extend the garden and add a fountain"), you can modify the text in the Positive Prompt node (<code>TextEncodeQwenImageEdit</code>).</li></ul></li><li>Run the Workflow:<ul><li>Queue the prompt in ComfyUI. The final image will be saved to your <code>ComfyUI/output/</code> folder.</li></ul></li></ol><hr /><h3 id="technical-details-and-nodes-used">Technical Details &amp; Nodes Used</h3>NodePurposeKey Settings<code>LoaderGGUF</code>Loads the main Qwen Image Edit model.<code>Qwen_Image_Edit-Q5_0.ggufClipLoaderGGUF</code>Loads the Qwen Vision encoder.<code>qwen2.5-vl-7b-it-q4_0.ggufVAELoader</code>Loads the Qwen VAE for encoding/decoding.<code>qwen_image_vae.safetensorsLoraLoaderModelOnly</code>Applies the Lightning LoRA for fast sampling.Strength: <code>1.0LoadImage</code>Loads your input image.<code>ImagePadForOutpaint</code>Core node. Adds transparent padding around the image for the AI to fill.Left/Right: <code>384</code>, Top/Bottom: <code>0</code>, Feather: <code>48ImageScaleToTotalPixels</code>Scales the padded image to an optimal size for the model.Megapixels: <code>0.9TextEncodeQwenImageEdit</code>Positive Prompt: Instructs the model on how to extend the image.<code>TextEncodeQwenImageEdit</code>Negative Prompt: Instructs the model on what to avoid (seams, artifacts).<code>VAEEncode</code>Encodes the scaled image into the latent space.<code>ModelSamplingAuraFlow</code>Configures the model for Aura Flow sampling.Shift: <code>3.0CFGNorm</code>Patches the model for CFG.Strength: <code>1.0KSampler</code>Performs the outpainting denoising process.Steps: <code>8</code>, Sampler: <code>euler</code>, CFG: <code>1.0</code>, Denoise: <code>0.95VAEDecode</code>Decodes the final latents back into an image.<code>SaveImage</code>Saves the finished, outpainted image.<hr /><h3 id="tips-for-best-results">Tips for Best Results</h3><ul><li>Image Choice: Start with images that have a clear and continuable background (e.g., skies, water, walls, fields). Results are most seamless when the AI has clear patterns to continue.</li><li>Prompt Guidance: The provided positive prompt is excellent for general use. For more creative control, try instructions like: "Extend the forest and add a path on the right," or "Continue the architecture in the same Gothic style."</li><li>Padding Settings: The <code>ImagePadForOutpaint</code> node is set to add <code>384</code> pixels to the left and right. You can adjust these values (e.g., <code>512</code>, <code>256</code>) to control how much the image is expanded in each direction.</li><li>Denoise Strength: The denoise value of <code>0.95</code> means the original image will be largely preserved. Lower values (e.g., <code>0.8</code>) will preserve it more but may be less creative; higher values give the AI more freedom but risk altering the original.</li></ul><hr /><h3 id="required-models-(download-links)">Required Models (Download Links)</h3><ol><li>Qwen_Image_Edit-Q5_0.gguf: <a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Qwen-Image-Edit-GGUF">https://huggingface.co/QuantStack/Qwen-Image-Edit-GGUF</a></li><li>qwen2.5-vl-7b-it-q4_0.gguf: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main/split_files/text_encoders">https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main/split_files/text_encoders</a></li><li>qwen_image_vae.safetensors: <a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Qwen-Image-Edit-GGUF">https://huggingface.co/QuantStack/Qwen-Image-Edit-GGUF</a></li><li>Qwen-Image-Edit-Lightning-8steps-V1.0-bf16.safetensors: <a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Qwen-Image-Lightning/tree/main">https://huggingface.co/lightx2v/Qwen-Image-Lightning/tree/main</a></li></ol><hr /><h3 id="conclusion">Conclusion</h3>This workflow showcases the incredible capability of the Qwen Image Edit model for outpainting tasks. It removes the technical barrier, providing a streamlined, one-click solution to creatively extend your images while maintaining perfect consistency. It's a must-try for photographers, digital artists, and anyone looking to expand their creative canvas.If you use this workflow, please share your results! I'd love to see what you create.

Files with SHA256: 9e6d7442e6512ce0b41ec130ace6af109ff22e88fdcfc5c2dc1a8fcad880ff45