Z-Image-Turbo/Base-AIO-Workflow

🌊 NEW WORKFLOW: Depth Anything V3 ControlNet EditionThis workflow introduces Depth Anything V3 - a cutting-edge depth estimation model that recovers visual geometry from any view with unprecedented accuracy.WHAT'S NEW:- Depth Anything V3 integration for state-of-the-art depth maps- Dual workflow modes: Generate images OR preview 3D point clouds- Toggle system for easy mode switching- Multi-view depth consistency (better than traditional methods)- Manual seed control for reproducibility- Megapixel scaling with automatic aspect ratioTWO POWERFUL MODES:1. GENERATE MODE (normalization: Standard/V2-Style): • Use depth for ControlNet-guided generation • Superior depth accuracy vs Canny/HED/traditional depth • Preserves scene composition and spatial relationships • ControlNet strength 0.6-0.8 recommended2. 3D PREVIEW MODE (normalization: Raw): • Visualize depth as interactive 3D point cloud • Export point clouds for external use • Perfect for understanding scene geometry • Great for pre-visualizationHOW IT WORKS:Upload image → DA3 extracts depth → Choose mode with toggles → Generate or Preview!CRITICAL TOGGLE SETTINGS:- For Generation: Toggle Generate ON, 3D Preview OFF- For 3D Preview: Toggle 3D Preview ON, Toggle Generate OFF⚠️ Don't mix modes - will cause errors!MODEL SIZES:- da3_small (80M) - Fast- da3_base (120M) - Good- da3_large (350M) - Recommended ⭐- da3_giant (1.15B) - Best quality8GB VRAM: Use up to da3_largeREQUIRED CUSTOM NODES:Depth Anything 3: <a target="_blank" rel="ugc" href="https://github.com/ByteDance-Seed/Depth-Anything-3">https://github.com/ByteDance-Seed/Depth-Anything-3</a>Same base settings: 9 steps, CFG 1.0, natural language promptsReplaces older ControlNet Union workflow - now focused on depth only!

zImageTurboBaseAIO_zitAIODepthv3.zip

ZIT-AIO-DepthV3

ZIB-AIO

ZIT-AIO-Anime

ZIT-AIO-SeedVR2

ZIT-AIO-Variance

ZIT-AIO-Control

ZIT-AIO-v2.0

ZIT-AIO-v1.0

Files with SHA256: 9dfa654d9978de5f8fa5a92af397f18d151dd90edf3587a39b74f3775300bebc