⚡ SWAPFACE PRO V1 — Ultimate ReActor Face Swap Pipeline&gt; The cleanest, most complete face-swap workflow for ComfyUI. Drop two images. Get cinematic results. No PhD required.---## 🎯 WHAT THIS WORKFLOW DOESSwapFace Pro V1 is a fully pre-configured ComfyUI workflow that swaps faces between two images using the ReActor extension with a triple-layer quality pipeline:- 🔁 Face Swap via <code>inswapper_128.onnx</code> — industry-standard accuracy- ✨ Dual GFPGANv1.4 Restoration — runs before AND after swap for maximum sharpness- 🎭 Smart Mask Blending via YOLOv8 + SAM — auto-feathered edges, zero harsh lines- 💾 Auto-saves to <code>SwapFace_Result_*.png</code> every run---## 🛠 REQUIRED MODELS- <code>inswapper_128.onnx</code> → <code>models/insightface/</code>- <code>GFPGANv1.4.pth</code> → <code>models/facerestore_models/</code>- <code>face_yolov8m.pt</code> → <code>models/ultralytics/bbox/</code>- <code>sam_vit_b_01ec64.pth</code> → <code>models/sams/</code>---## 📦 REQUIRED EXTENSIONInstall ComfyUI-ReActor by Gourieff via ComfyUI Manager.Also install: <code>onnxruntime-gpu</code>, <code>insightface</code>, <code>segment_anything</code>---## 🚀 HOW TO USE1. Load your source face image in the SOURCE FACE node2. Load your target image in the TARGET IMAGE node3. Click Queue Prompt4. Done — result appears in FINAL RESULT + saved to output folder---## 💡 PRO TIPS- Use high-res, front-facing source faces for best accuracy- Increase blur_radius to 14–18 in MaskHelper for soft outdoor blending- Use input_faces_index to target specific faces in group photos- Match lighting between source and target for seamless results---&gt; ⭐ If this saved you hours of setup — leave a Like and share your results in the comments! Every piece of feedback helps improve future versions.

SWAPFACE PRO V1

## 📋 VERSION NOTES### V1.0 — Initial Release- ✅ Full ReActor face-swap pipeline (inswapper_128.onnx)- ✅ Dual GFPGANv1.4 restoration (FaceBoost + main swap)- ✅ Smart mask blending via YOLOv8 + SAM- ✅ Live mask preview node- ✅ Auto-save output (SwapFace_Result_*.png)- ✅ All parameters pre-tuned — zero config required---### 🔜 PLANNED FOR V2- [ ] Multi-face batch support- [ ] ControlNet integration for pose-aware swapping- [ ] Custom mask input option- [ ] Support for video frame sequences- [ ] IP-Adapter face consistency mode---&gt; 💬 Have a feature request? Drop it in the comments — top voted ideas go straight into V2.

Files with SHA256: 8c14daf3044c7a85af5d1e7e5d749518334c9356aa08ae30875c36801e4ae389