<h2 id="comfyui-flux-continuum-modular-interface-u14rx66t9">ComfyUI Flux Continuum - Modular Interface</h2>A modular workflow that brings order to the chaos of image generation pipelines.📺 <a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=cjWuPcRZ1j0">Watch the Tutorial</a>🔗 <a target="_blank" rel="ugc" href="https://github.com/robertvoy/ComfyUI-Flux-Continuum">GitHub</a><h2 id="updates-8dv12ihg2">Updates</h2><ul><li>1.7.0: Enhanced workflow and usability update 📺 <a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=e_7cYbBwjFc">Watch Video Update</a><ul><li>Image Transfer Shortcut: Use <code>Ctrl+Shift+C</code> to copy images from Img Preview to Img Load (customizable in Settings &gt; Keybinding &gt; Image Transfer)</li><li>Configurable Model Router: Dynamic model selection with customizable JSON mapping for flexible workflows</li><li>Hint System: Interactive hint nodes provide contextual help throughout the workflow</li><li>Crop &amp; Stitch: Enhanced inpainting/outpainting with automatic crop and stitch functionality</li><li>Smart Guidance: Automatic guidance value of 30 for inpainting, outpainting, canny, and depth operations</li><li>TeaCache Integration: Optional speed boost for all outputs (trades some quality for performance)</li><li>Improved Preprocessor Preview Logic: CN Input is used for previewing when ControlNet strength &gt; 0, otherwise uses Img Load</li><li>Workflow Reorganization: Modules reordered for more logical flow</li><li>Redux Naming: IP Adapter renamed to Redux for consistency with BFL terminology</li></ul></li></ul><h2 id="overview-e9drnd48l">Overview</h2>ComfyUI Flux Continuum revolutionises workflow management through a thoughtful dual-interface design:<ul><li>Front-end: A consistent control interface shared across all modules</li><li>Back-end: Powerful, modular architecture for customisation and experimentation</li></ul><h2 id="core-features-bhiz2n71t">✨ Core Features</h2>Perfect for creators who want a consistent, streamlined experience across all image generation tasks, while maintaining the power to customize when needed.<ul><li>Unified Control Interface<ul><li>Single set of controls affects all relevant modules</li><li>Smart guidance adjustment based on operation type</li><li>Consistent experience across all generation tasks</li></ul></li><li>Smart Workflow Management<ul><li>Only activates nodes and models required for current task</li><li>Toggle between different output types seamlessly</li><li>Efficiently handles resource allocation</li><li>Optional TeaCache for speed optimization</li></ul></li><li>Universal Model Integration<ul><li>LoRAs, ControlNets and Redux work across all output modules</li><li>Seamless Black Forest Labs model support</li><li>Configurable model routing for custom workflows</li></ul></li><li>Enhanced Usability<ul><li>Interactive hint system for contextual help</li><li>Quick image transfer with keyboard shortcut</li><li>Intelligent preprocessing based on control values</li><li>Crop &amp; stitch for seamless inpainting/outpainting</li></ul></li></ul><hr /><h2 id="quick-start-sj57ft9re">🚀 Quick Start</h2>📺 New to Flux Continuum? <a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=cjWuPcRZ1j0">Watch the tutorial first</a><ol><li>Clone repo to the custom nodes folder</li></ol><pre><code>git clone https://github.com/robertvoy/ComfyUI-Flux-Continuum</code></pre><ol><li><a target="_blank" rel="ugc" href="https://github.com/robertvoy/ComfyUI-Flux-Continuum/blob/main/workflow/Flux%2B%201.7.0_release.json">Download</a> and import the workflow into ComfyUI</li><li>Install missing custom nodes using the ComfyUI Manager</li><li>Configure your models in the config panel (press <code>2</code> to access)</li><li>Download any missing models (see Model Downloads section below)</li><li>Return to the main interface (press <code>1</code>)</li><li>Select <code>txt2img</code> from the output selector (top left corner)</li><li>Run the workflow to generate your first image</li></ol><hr /><h2 id="usage-guide-iofe91s4j">🎯 Usage Guide</h2><h3 id="output-selection-n85s3fl87">Output Selection</h3>The workflow is controlled by the Output selector in the top-left corner. Select your desired output and all relevant controls will automatically apply.<h3 id="key-controls-d0emzwy5s">Key Controls</h3>🎨 Main Generation<ul><li>Prompt: Your text description for generation</li><li>Denoise: Controls strength for img2img operations (0 = no change, 1 = completely new)</li><li>Steps: Number of sampling steps (higher = more detail, slower)</li><li>Guidance: How closely to follow the prompt (automatically set to 30 for inpainting/outpainting/canny/depth)</li><li>TeaCache: Toggle for speed boost (some quality trade-off)</li></ul>🖼️ Input Images<ul><li>Img Load: Primary image for all img2img operations (inpainting, outpainting, detailer, upscaling)</li><li>CN Input: Source for ControlNet preprocessing</li><li>Redux 1-3: Up to 3 reference images for style transfer (use very low strength values)</li><li>Tip: Use <code>Ctrl+Shift+C</code> to quickly copy from Img Preview to Img Load</li></ul>🎛️ ControlNet &amp; Redux<ul><li>ControlNets activate when strength &gt; 0</li><li>When CN strength &gt; 0, preprocessor uses CN Input; otherwise uses Img Load</li><li>Preview preprocessor results by selecting corresponding output (e.g., "preprocessor canny")</li><li>Redux sliders control each Redux input individually (1 = Redux 1, etc.)</li></ul>Recommended ControlNet Values:<ul><li>Canny: Strength=0.7, End=0.8</li><li>Depth: Strength=0.8, End=0.8</li><li>Pose: Strength=0.9, End=0.65</li></ul>🔧 Image Processing<ul><li>Resize, crop, sharpen, color correct, or pad images</li><li>Preview results with "imgload prep" output</li><li>Bypass nodes after processing to avoid reprocessing (<code>Ctrl+B</code>)</li></ul>⬆️ Upscaling<ul><li>Resolution Multiply: Multiplies image resolution after any preprocessing</li><li>Upscale Model: Choose your upscaling model (recommended: 4xNomos8kDAT)</li><li>📺 <a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=TmF3JK_1AAs">Watch Upscaling Tutorial</a></li></ul><hr /><h2 id="model-downloads-xpik3o6hh">📥 Model Downloads</h2><h3 id="required-models-4lb3idxwi">Required Models</h3>unet folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-dev/resolve/main/flux1-dev.safetensors">flux1-dev.safetensors</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-Depth-dev/resolve/main/flux1-depth-dev.safetensors">flux1-depth-dev.safetensors</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-Canny-dev/resolve/main/flux1-canny-dev.safetensors">flux1-canny-dev.safetensors</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-Fill-dev/resolve/main/flux1-fill-dev.safetensors">flux1-fill-dev.safetensors</a></li></ul>Note: If you don't use Canny or Depth models, you can bypass their load nodes and skip downloading them.vae folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-dev/resolve/main/ae.safetensors">ae.safetensors</a></li></ul>clip folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp8_e4m3fn.safetensors">t5xxl_fp8_e4m3fn.safetensors</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors">clip_l.safetensors</a></li></ul>style_models folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.1-Redux-dev/resolve/main/flux1-redux-dev.safetensors">flux1-redux-dev.safetensors</a></li></ul>clip_vision folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/sigclip_vision_384/resolve/main/sigclip_vision_patch14_384.safetensors">sigclip_vision_patch14_384.safetensors</a></li></ul>controlnet/FLUX folder:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Shakker-Labs/FLUX.1-dev-ControlNet-Union-Pro-2.0/resolve/main/diffusion_pytorch_model.safetensors">FLUX.1-dev-ControlNet-Union-Pro-2.0.safetensors</a> (rename file)</li></ul>

FLUX Continuum (Modular Interface for ComfyUI)

1.6.4: ControlNet Union Pro v2 Update 📺 <a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=oh1P_4d9_HI">Watch Video Update</a><ul><li>ControlNet Union Pro v2: Integrated the new Depth, Canny, OpenPose ControlNets</li><li>New canny preprocessor control</li><li>Removed the inputs preview tab</li><li>Better upscaling controls</li><li>New Redux (IPAdapter) implementation</li></ul>

Files with SHA256: ae46a428bed8d2906b288cc27663996dede429656a39bbec702355270bdcce10