Where To Go Next
After you understand the basic workflow, choose the next topic based on what you are trying to do. Diffusion has many branches, and learning everything at once usually makes results harder to control.
Choose By Problem, Not By Feature List
The next page to read depends on what you are trying to improve: tool choice, prompt control, workflow control, model selection, troubleshooting, or deeper theory.
If You Need To Choose A Tool
Go to Generation Tools when you want to compare local interfaces, package managers, and hosted generation websites.
Start with the kind of workflow you want, not the longest feature list. A simple hosted generator may be enough for prompt practice, while local tools make more sense when you need custom models, LoRAs, privacy, or repeatable workflows.
If You Want Better Prompts
Go to prompting when you want to improve subject control, composition, style consistency, or negative prompt use.
Useful topics to learn:
- prompt structure
- prompt weighting
- negative prompts
- model-specific prompting habits
- how prompts interact with LoRAs and other conditioning
If You Want Better Control
Go to workflows when text prompting alone is not enough.
Useful topics to learn:
- image-to-image
- inpainting
- upscaling and refinement
- pose, depth, edge, and segmentation guidance
- multi-stage refinement
If You Want A Specific Style Or Subject
Go to LoRAs and model pages when you need a model or adapter to provide knowledge the base model does not reliably contain.
Useful topics to learn:
- checkpoint selection
- LoRA compatibility
- LoRA weights and trigger words
- embeddings
- model architecture compatibility
If Images Look Wrong
Go to troubleshooting when you have a visible symptom.
Common starting points:
- distorted hands, faces, or text
- muddy detail or oversharpening
- prompt being ignored
- repeated objects
- color or contrast problems
- out-of-memory errors
Troubleshooting works best when you can describe the symptom, the model, the workflow, and the settings that produced it.
If You Want To Understand The System
Go to theory when you want a deeper explanation of denoising, latent diffusion, text encoders, guidance, model architecture, or training.
Theory is useful, but you do not need to read it all before making images. Return to it when practical problems make you curious about why a setting behaves the way it does.
Learning Path
You can use this site out of order. A practical path is to start with one simple workflow, learn the terms that appear in your UI, then branch into prompting, LoRAs, troubleshooting, or theory as real questions come up.