<h1 id="z-image-turbo-advanced-workflows">🚀 Z-Image-Turbo Advanced Workflows</h1>By <a rel="ugc" href="https://civarchive.com/user/sphiratrioth666">sphiratrioth666</a>Enhanced workflows with sliders, multi-LoRA, VRAM management, and professional detailing.work - these are serious upgrades! 🚀<hr /><h3 id="advanced-workflows-by-sphiratrioth666">🎉 Advanced Workflows by sphiratrioth666 </h3><h3 id="these-workflows-were-modified-and-enhanced-by-user-sphiratrioth666-based-on-the-original-z-image-turbo-aio-workflows.">These workflows were modified and enhanced by user <a rel="ugc" href="https://civarchive.com/user/sphiratrioth666">sphiratrioth666</a>, based on the original <a rel="ugc" href="https://civarchive.com/models/2174008?modelVersionId=2478180">Z-Image-Turbo-AIO</a> workflows. </h3><h3 id="he-integrated-meaningful-improvements-and-created-an-entirely-new-detailer-workflow-with-sam2-+-segs!">He integrated meaningful improvements and created an entirely new DETAILER workflow with SAM2 + SEGS! </h3><h3 id="big-thanks-and-full-credits-go-to-him!">Big thanks and full credits go to him! 🙏 </h3><h3 id="check-out-his-work-these-are-serious-upgrades!">Check out his work - these are serious upgrades! 🚀</h3><hr /><h2 id="three-workflow-variants:">📦 Three Workflow Variants:</h2><h3 id="1.-txt2img-(advanced)">1. 🎨 Txt2Img (Advanced)</h3>Pure text-to-image with advanced controlsKey Features:<ul><li>Interactive sliders (CFG, Steps, Denoise, Upscale)</li><li>4 LoRA slots with individual strength controls</li><li>PURGE VRAM automatic cleanup</li><li>Play/Stop regeneration system (save only finals!)</li><li>2D resolution slider (3:4, 4:3, 16:9)</li><li>Improved preview/save order (LQ → HQ)</li></ul>Use when: You want maximum control over text-to-image generationExtra nodes: MXToolkit, LayerUtility<hr /><h3 id="2.-img2img-+-controlnet">2. 🎮 Img2Img + ControlNet </h3>ControlNet precision with advanced controlsKey Features:<ul><li>All Txt2Img features PLUS:</li><li>ControlNet Union (Canny, Depth, Pose, HED, MLSD)</li><li>Megapixel scaling (auto aspect ratio)</li><li>ControlNet strength slider</li><li>Input image guidance</li></ul>Use when: You need precise control with reference imagesExtra downloads: ControlNet Union file (~2.5GB) Save in: <code>ComfyUI/models/model_patches/</code><hr /><h3 id="3.-img2img-+-controlnet-+-detection">3. ✨ Img2Img + ControlNet + DETECTION</h3>Professional pipeline with selective detailingKey Features:<ul><li>Everything from Img2Img PLUS:</li><li>SAM2 auto-segmentation (face, hands, details)</li><li>SEGS selective detailer</li><li>Model upscaler (4x to 10K resolution)</li><li>Grain addition for photographic look</li><li>Smart resize to 2K/4K (reasonable file size)</li><li>Multiple PURGE VRAM nodes</li></ul>Use when: You need maximum quality for final outputsExtra nodes: Impact Pack, SAM2 Extra downloads: Upscaler model, SAM2 model<hr /><h2 id="quick-comparison:">🆚 Quick Comparison:</h2>| Feature | Txt2Img | Img2Img | Detailer ||---------|---------|---------|----------|| Input Image | ❌ | ✅ | ✅ || ControlNet | ❌ | ✅ | ✅ || SEGS Detailer | ❌ | ❌ | ✅ || Complexity | Simple | Medium | Advanced || Speed | Fast (3-5s) | Medium (5-10s) | Slow (20-60s) || Quality | High | Higher | Maximum || Use Case | Quick gens | Controlled gens | Final portfolio |<hr /><h2 id="when-to-use-which:">🎯 When to Use Which:</h2><h3 id="choose-txt2img-when:">Choose Txt2Img when:</h3>✅ Pure text-to-image generation ✅ Quick iterations and testing ✅ Multiple LoRAs experimentation ✅ Don't need reference images<h3 id="choose-img2img-+-controlnet-when:">Choose Img2Img + ControlNet when:</h3>✅ Have reference/input image ✅ Need pose/composition control ✅ Sketch-to-photo conversion ✅ Architectural work ✅ Want guided generation<h3 id="choose-detailer-when:">Choose Detailer when:</h3>✅ Creating portfolio pieces ✅ Professional/commercial work ✅ Need perfect face/hand details ✅ Want photorealistic texture ✅ Maximum quality required ✅ Don't mind longer processing<hr /><h2 id="shared-features-(all-3):">✨ Shared Features (All 3):</h2>🎛️ Interactive Sliders:<ul><li>CFG, Steps, Denoise</li><li>LoRA strengths (4 slots)</li><li>Upscale parameters</li><li>ControlNet strength (Img2Img variants)</li></ul>🔄 Play/Stop System:<ul><li>Green PLAY = Generate/regenerate</li><li>Purple SAVE = Save final only</li><li>No cluttered saves folder!</li></ul>🧹 PURGE VRAM:<ul><li>Automatic cleanup after generation</li><li>Prevents memory buildup</li><li>Better performance on all GPUs</li></ul>📦 Multi-LoRA:<ul><li>4 LoRA slots</li><li>Individual strength sliders</li><li>Easy on/off (set to 0.0)</li></ul>📸 Metadata:<ul><li>Auto-saved to images</li><li>Easy CivitAI uploads</li></ul><hr /><h2 id="downloads:">📥 Downloads:</h2>Main Model: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2173571?modelVersionId=2448013">Z-Image-Turbo-AIO FP8/BF16</a>ControlNet Union (for Img2Img variants): <a target="_blank" rel="ugc" href="https://huggingface.co/alibaba-pai/Z-Image-Turbo-Fun-Controlnet-Union/resolve/main/Z-Image-Turbo-Fun-Controlnet-Union.safetensors?download=true">HuggingFace Download</a> ⚠️ Save in: <code>ComfyUI/models/model_patches/</code>Test Online: <a target="_blank" rel="ugc" href="https://tensorart.me/de-DE/models/939302608026244936/Z-Image-Turbo-AIO-FP8-FP8">TensorArt (FP8)</a><hr /><h2 id="required-custom-nodes:">🎯 Required Custom Nodes:</h2>All Workflows:<ul><li><a target="_blank" rel="ugc" href="https://github.com/komojini/ComfyUI_MXKIT">MXToolkit</a> - Sliders &amp; controls</li><li><a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">rgthree-comfy</a> - LoRA stack</li><li><a target="_blank" rel="ugc" href="https://github.com/chflame163/ComfyUI_LayerStyle">LayerUtility</a> - PURGE VRAM</li><li><a target="_blank" rel="ugc" href="https://github.com/edelvarden/comfyui_image_metadata_extension">comfyui_image_metadata_extension</a></li></ul>Img2Img + ControlNet:<ul><li><a target="_blank" rel="ugc" href="https://github.com/Fannovel16/comfyui_controlnet_aux">comfyui_controlnet_aux</a> - Preprocessors</li><li>⚠️ ComfyUI 3.77+ required!</li></ul>Detailer:<ul><li><a target="_blank" rel="ugc" href="https://github.com/ltdrdata/ComfyUI-Impact-Pack">Impact Pack</a> - SEGS detailer</li><li><a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-segment-anything-2">SAM2</a> - Segmentation</li><li>⚠️ ComfyUI 3.77+ required!</li></ul><hr /><h2 id="settings-(all-workflows):">⚙️ Settings (All Workflows):</h2>Steps: 9 (slider adjustable) CFG: 1.0 (slider adjustable) Sampler: res_multistep or euler_ancestral Scheduler: simple or beta NO negative prompts needed<hr /><h2 id="pro-tips:">💡 Pro Tips:</h2>Slider Workflow:<ul><li>Start with defaults, adjust as needed</li><li>Set LoRA to 0.0 to disable</li><li>Use PLAY to test variations</li><li>Only SAVE final results</li></ul>ControlNet Strength:<ul><li>0.3-0.5 = Subtle guidance</li><li>0.6-0.8 = Balanced (recommended)</li><li>0.9-1.0 = Strong control</li></ul>Detailer:<ul><li>Best with 1024px+ input</li><li>Let SAM2 auto-detect regions</li><li>Grain at 10K = most natural</li><li>Downscale to 2K/4K recommended</li></ul>PURGE VRAM:<ul><li>Runs automatically</li><li>Helps weaker GPUs</li><li>Prevents memory issues</li></ul><hr /><h2 id="example-workflow:">🎨 Example Workflow:</h2>Quick Test (Txt2Img):<ol><li>Load 1-2 LoRAs → 2. Write prompt → 3. PLAY → 4. Adjust sliders → 5. PLAY again → 6. SAVE</li></ol>Controlled Gen (Img2Img):<ol><li>Upload reference → 2. Choose preprocessor → 3. Load LoRAs → 4. Write prompt → 5. Adjust strength → 6. PLAY → 7. SAVE</li></ol>Final Polish (Detailer):<ol><li>Upload input → 2. Set up ControlNet → 3. Load LoRAs → 4. Write prompt → 5. PLAY (wait 30-60s) → 6. SAVE 2K/4K</li></ol><hr /><h2 id="faq:">❓ FAQ:</h2>Q: Which workflow should I start with? A: Txt2Img for learning, Img2Img for control, Detailer for finals.Q: Do I need all custom nodes for all workflows? A: No - each workflow lists its specific requirements.Q: What's MXToolkit? A: Provides the slider interface. Makes adjustments easier.Q: Why PURGE VRAM? A: Cleans memory after generation. Especially helpful on 8GB cards.Q: Detailer too slow? A: Yes, it's intensive. Use only for final images, not testing.Q: Can I use original workflows instead? A: Yes! These are advanced versions. Original workflows still work great.Q: MXToolkit not working with ComfyUI 2.0? A: Disable Node 2.0 interface for now. MXToolkit compatibility coming.<hr /><h2 id="credits:">🙏 Credits:</h2>Advanced Workflows: <a target="_blank" rel="ugc" href="https://civarchive.com/user/sphiratrioth666">sphiratrioth666</a> Original Workflows: <a rel="ugc" href="https://civarchive.com/user/SeeSeeLP">SeeSeeLP</a> Base Model: Tongyi Lab (Alibaba Group) - Z-Image-Turbo License: Apache 2.0Big thanks to sphiratrioth666 for the amazing enhancements! 🎉<hr /><h2 id="system-requirements:">📊 System Requirements:</h2>Minimum:<ul><li>VRAM: 8GB (all workflows tested on RTX 4060)</li><li>RAM: 16GB (32GB recommended for Detailer)</li><li>ComfyUI: 3.77+ (for ControlNet/Detailer)</li></ul>Detailer additionally needs:<ul><li>More processing time (~30-60s)</li><li>SAM2 and upscaler models</li><li>Patience! 😄</li></ul><hr />Updated: December 2025 Compatible: Z-Image-Turbo-AIO FP8 &amp; BF16 Tested: RTX 4060 8GB, RTX 5090<hr />"I upgraded your Z-Image workflow by a lot" - sphiratrioth666Try all three workflows and find your perfect setup! 🚀

Z-Image-Turbo-AIO-Workflow

Z-Image-Turbo Img2Img Detection - By sphiratrioth666Professional workflow with SAM2, SEGS detailing, model upscaler, and grain🎯 Key Features:✅ SAM2 Segmentation - Precise area detection ✅ SEGS Detailer - Selective detail enhancement ✅ Model Upscaler - High-quality upscaling ✅ Grain Addition - Natural photographic look ✅ Smart Resize - Downscale to 2K/4K for reasonable file size ✅ Multi-LoRA Support - 4 LoRA slots ✅ PURGE VRAM - Extensive cleanup ✅ Play/Stop System - Test before saving📦 Required Downloads:Models:<ul><li><a target="_blank" rel="ugc" href="https://civitai.com/models/2173571?modelVersionId=2448013">Z-Image-Turbo-AIO FP8/BF16</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/alibaba-pai/Z-Image-Turbo-Fun-Controlnet-Union/resolve/main/Z-Image-Turbo-Fun-Controlnet-Union.safetensors?download=true">Z-Image-Turbo-Fun-Controlnet-Union.safetensors</a></li><li>Model Upscaler (any 4x upscaler like ESRGAN)</li><li>SAM2 Model - For segmentation</li><li>⚠️ ControlNet in: <code>ComfyUI/models/model_patches/</code></li></ul>Custom Nodes:<ul><li><a target="_blank" rel="ugc" href="https://github.com/komojini/ComfyUI_MXKIT">MXToolkit</a> - Sliders</li><li><a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">rgthree-comfy</a> - LoRA stack</li><li><a target="_blank" rel="ugc" href="https://github.com/Fannovel16/comfyui_controlnet_aux">comfyui_controlnet_aux</a></li><li><a target="_blank" rel="ugc" href="https://github.com/chflame163/ComfyUI_LayerStyle">LayerUtility</a> - PURGE + effects</li><li><a target="_blank" rel="ugc" href="https://github.com/ltdrdata/ComfyUI-Impact-Pack">Impact Pack</a> - SEGS detailer</li><li><a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-segment-anything-2">SAM2</a> - Segmentation</li><li><a target="_blank" rel="ugc" href="https://github.com/edelvarden/comfyui_image_metadata_extension">comfyui_image_metadata_extension</a></li></ul>⚠️ ComfyUI 3.77+ required!⚙️ Settings:Steps: 9 (adjustable) CFG: 1.0 (adjustable) Sampler: res_multistep or euler_ancestral Scheduler: simple or beta Guide Size: Adjustable (SAM2) Max Size: Adjustable (SEGS)🎨 Workflow Pipeline:<ol><li>Input Image → ControlNet guidance</li><li>SAM2 Segmentation → Auto-detect regions</li><li>SEGS Detailer → Enhance face, hands, details</li><li>Model Upscale → 4x upscaling (10K resolution)</li><li>Add Grain → Natural photographic texture</li><li>Resize to 2K/4K → Reasonable file size (~10-20MB)</li></ol>🔍 How to Use:<ol><li>Upload input image</li><li>Choose preprocessor - ControlNet type</li><li>Load LoRAs - Up to 4 slots</li><li>Write prompt - Detailed description</li><li>Adjust sliders - CFG, Steps, Guide/Max size</li><li>Click PLAY - Full detailing pipeline runs</li><li>Review result - Check detailing quality</li><li>Click SAVE - Final 2K/4K output</li></ol>🎯 What Makes It Special:SAM2 Segmentation:Automatically detects faces, hands, and important areas for selective enhancement.SEGS Detailer:Applies focused detail enhancement only where needed. Fixes common issues like:<ul><li>Blurry faces</li><li>Unclear hands</li><li>Lost details in upscaling</li></ul>Model Upscaler:Uses high-quality model upscaler (ESRGAN-style) to reach 10K resolution.Grain Addition:Adds subtle photographic grain at 10K resolution for natural look. Blends enhanced areas seamlessly.Smart Resize:Downscales back to 2K (2048px) or 4K (4096px) for:<ul><li>Reasonable file sizes (10-20MB vs 100MB+)</li><li>Easier sharing and storage</li><li>Still exceptional quality</li></ul>🎛️ Slider Controls:<ul><li>Guide Size - SAM2 detection sensitivity</li><li>Max Size - Maximum detail patch size</li><li>CFG - Guidance strength</li><li>Steps - Generation steps</li><li>Denoise - Detail enhancement strength</li><li>LoRA 1-4 - Individual LoRA strengths</li></ul>💡 Advanced Tips:For Best Results:<ul><li>Use high-quality input (1024px+)</li><li>Let SAM2 auto-detect regions</li><li>Adjust Guide Size if detection is off</li><li>Higher Denoise = more detail changes</li><li>Grain at 10K resolution looks most natural</li></ul>When to Use:<ul><li>Final generations for portfolio</li><li>Professional work requiring max quality</li><li>Images with faces/hands needing detail</li><li>When you want photorealistic texture</li><li>Publication-quality outputs</li></ul>Performance:<ul><li>Takes longer than simple workflows (~20-60 seconds)</li><li>Uses more VRAM (PURGE helps!)</li><li>Worth it for final results</li></ul>🔧 Downscale/BYPASS:Workflow includes option to:<ul><li>Downscale - Resize to 2K/4K (recommended)</li><li>BYPASS - Keep full 10K resolution (huge files!)</li></ul>Most users prefer downscaled version for balance of quality and file size.🧹 PURGE VRAM:Multiple PURGE nodes throughout pipeline prevent VRAM buildup during complex processing.Credits: Advanced detailer workflow by <a target="_blank" rel="ugc" href="https://civitai.com/user/sphiratrioth666">sphiratrioth666</a> Based on: Z-Image-Turbo ControlNet workflow by SeeSeeLP Version: Advanced NQ + ControlNet + SEGS Detailer Edition Updated: December 2024For professionals who need maximum quality and detail! ✨"A bit over the top, but it works very well" - sphiratrioth666

Files with SHA256: 9cc520e6d0460ff9d4069948b3b393c3d9b75f0242e246fe4dcdded31fe25713