ComfyUI & Stable Diffusion: Setup, Models, and Workflows
16 posts in this series
This series connects ComfyUI basics with real Stable Diffusion decisions. Start here if you want to install ComfyUI, make shared workflows reproducible, choose between SDXL, SD 3.5, and FLUX, or build prompt templates you can actually reuse.
ComfyUI Beginner Guide: Install It, Read the Node UI, Place Models, and Generate Your First Image
A practical ComfyUI beginner guide covering Desktop, Portable, Manual, and Cloud setup, model folders, the default node workflow, and first-run troubleshooting.
ComfyUI Workflow Reuse Guide: A Troubleshooting Checklist from Import to Reproduction
Red nodes, empty model dropdowns, or results that do not match the sample image after importing a ComfyUI workflow? This guide walks through JSON and PNG metadata import, custom node fixes, model path mapping, and reproducible workflow packaging.
Stable Diffusion Model Selection Guide: Practical Decisions from Image Quality to Licensing
Not sure which Stable Diffusion model to choose? This guide compares SDXL, SD 3.5, and FLUX.1 by image quality, VRAM, ecosystem maturity, and commercial licensing, with ComfyUI test steps and common pitfalls.
Stable Diffusion Prompt Template Guide: Product Photos, Avatars, Posters, and Game Assets
Don't know how to write Stable Diffusion prompts? This guide breaks down positive prompts, negative prompts, weight syntax, ComfyUI iteration, and commercial license checks across four practical scenarios: product photos, avatars, posters, and game assets.
ComfyUI ControlNet: Control Poses and Composition with OpenPose, Canny, and Depth
Set up ControlNet in ComfyUI, match preprocessors to models, tune strength and sampling ranges, combine controls, and troubleshoot weak or missing effects.
Using LoRA in ComfyUI: Loading, Weights, Stacking, and Character Consistency
Load LoRA files in ComfyUI, tune strength_model and strength_clip, stack multiple LoRAs, match the base model, reduce character drift, and troubleshoot weak results.
ComfyUI IPAdapter: Reference Image Style, Composition, and Model Pairing
Pair IPAdapter with the right CLIP Vision model, tune weight and weight_type, control style and composition, and troubleshoot SD1.5, SDXL, and FLUX workflows.
ComfyUI FLUX.1 Kontext: Multi-Turn Editing and Character Consistency
Set up FLUX.1 Kontext Dev in ComfyUI, write preservation prompts, manage multi-turn edits, roll back drift, and choose the right consistency tool.
ComfyUI Face Consistency: InstantID vs FaceID vs ReActor
Compare InstantID, IPAdapter FaceID, and ReActor by workflow stage and model setup, with reference checks, tuning, troubleshooting, and licensing limits.
Train Your Own LoRA with ComfyUI: Data, Captions, Training, and Validation
Prepare a dataset, write captions and trigger words, tune dim, alpha, learning rates, steps, and VRAM settings, then diagnose fit and validate the LoRA in ComfyUI.
ComfyUI Regional Prompting for Multi-Subject Compositions: Regional, Cutoff, and Conditioning Combine
Separate ComfyUI subjects with Set Area, RegionalPrompt, and Cutoff, covering coordinates, mask sampling, ControlNet pairing, strength, and second-pass color bleed.
ComfyUI Upscaling and Inpainting: Hires Fix, FaceDetailer, and Local Repairs
Choose between pixel upscaling, latent resampling, FaceDetailer, manual inpainting, and tiled upscaling, then troubleshoot face drift, seams, and mask spill.
ComfyUI Low-VRAM Optimization: Run SDXL, FLUX, and Video Workflows on 6–8GB GPUs
Run ComfyUI on 6–8GB GPUs by budgeting model weights, T5, VAE, resolution, and batch size, then troubleshoot OOM errors and slow FLUX or video workflows.
ComfyUI Video Generation: Wan Text-to-Video, Image-to-Video, and AnimateDiff
Run your first ComfyUI video workflow with Wan or AnimateDiff, covering model folders, custom nodes, frames, FPS, VRAM limits, export, and troubleshooting.
Automate Batch Image Generation with the ComfyUI API
Export a ComfyUI API-format workflow, submit it through /prompt, wait over WebSocket, download outputs, and add batching, queue limits, and backend job tracking.
ComfyUI Troubleshooting: Red Nodes, VAE Errors, and Rollbacks
Diagnose red nodes, Import failed errors, loading screens, gray or black VAE output, and update conflicts in ComfyUI with a symptom-first workflow.
