YouTube: https://youtu.be/sejIPfdMNv8 Happy New Year! 🎉 In this video, we test the updated Qwen Image Edit v2511 and answer the big question: is it actually better than v2509? You’ll see side-by-side comparisons in ComfyUI (same seed, same CLIP/VAE) to evaluate consistency, identity preservation, and overall image quality. We also break down the real-world differences between model formats and sizes—including BF16, FP8, FP8 Lightning, GGUF (Q6/Q4), and Lightning LoRA—so you’ll know what matters for quality, speed, and VRAM. Finally, we explore sampler + scheduler combos (including Flow Match Euler Discrete Scheduler) to show how much they can change composition, depth, and lighting—even with the same prompt. ✅ All workflows shown are attached.

YouTube: https://youtu.be/EP8zjuxYjWE Merry Christmas! 🎄 In this video I test the brand-new Qwen Image Edit 2511 in ComfyUI and show what’s improved—especially the fact that popular LoRAs (including relighting) are now built into the base model, so you can relight images without adding a separate LoRA. That said, 2511 still has a major issue: image drifting (objects shift and details change). I’ll show you a relighting workflow workaround that locks the subject in place, supports very large inputs, and finishes with Frequency Detail Restore to bring back crisp original detail while keeping the new lighting. I've attached the basic workflow. Want the premium workflow?—Join the community to download it: https://myaiforce.com/pg Happy holidays, and enjoy the workflow!

Hey everyone! I'm excited to share a workflow that allows you to easily remove objects/people by painting a mask over them. You can find the model download link in the notes of the workflow. If you're running low on VRAM, don’t worry! You can also use the GGUF versions of the model. This workflow maintains image quality because it only resamples the specific area where you want the object removed, then seamlessly integrates the resampled image back into the original. It's a more efficient and faster option compared to Qwen Edit/Flux Kontext! And don’t forget to subscribe to my YouTube channel for more insights and tutorials on ComfyUI: https://www.youtube.com/@my-ai-force

YouTube Tutorial: https://youtu.be/HDVh4rYC9OQ Uploaded Models --- 1_Set-Model --- 🌟Load Diffusion Model🌟 - Download Link: https://civitai.com/models/141592/pixelwave - File Names: pixelwave_flux1_dev_fp8_03.safetensors - Directory: ComfyUI/models/unet/ 🌟DualCLIPLoader🌟 - Download Link: https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main - File Names: clip_l.safetensors & t5xxl_fp8_e4m3fn.safetensors - Directory: ComfyUI/models/clip/ 🌟 Load VAE 🌟 - Download Link: https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/ae.safetensors - File Name: ae.safetensors - Directory: ComfyUI/models/vae/ 🌟Load StyIe Model🌟 - Download Link: https://huggingface.co/black-forest-labs/FLUX.1-Redux-dev/tree/main - File Name: flux1-redux-dev.safetensors - Directory: ComfyUI/models/style_models/ 🌟Load Clip Vision🌟 - Download Link: https://huggingface.co/Comfy-Org/sigclip_vision_384/tree/main - File Name: sigclip_vision_patch14_384.safetensors - Directory: ComfyUI/models/clip-vision/ --- 2_Set-Layout --- 🌟DetectFaces🌟 - Download Link: https://huggingface.co/bluefoxcreation/FaceAlignment/tree/main - File Name: fan2_68_landmark.onnx - Directory: ComfyUI/models/landmarks --- 4_Hand-Detailer --- 🌟UltralyticsDetectorProvider🌟 - Download Link: https://huggingface.co/Bingsu/adetailer/tree/main - File Name: hand_yolov9c.pt - Directory: ComfyUI/models/ultralytics/ --- 7_PuLID-FaceSwap --- 🌟 Load PuLID Flux Model🌟

YouTube: https://youtu.be/pYqMQwBbECk Learn how to generate massive, ultra-high-resolution images using the Z Image Turbo model in ComfyUI. In this tutorial, we break down the full workflow—from the initial text-to-image step, to intelligent tiling, SeedVR2 upscaling, and a final detail-enhancement pass with Ultimate SD Upscale. You’ll see how splitting the image into tiles prevents out-of-memory issues, how each group in the workflow contributes to the final 8832×12192 result, and how to get sharper textures using SkinDiffDetail. Join our paid group to get the workflow: https://myaiforce.com/pg