Start here

The concept pages in the order a newcomer needs them, with the lesson that puts each group into practice.

#1. What you are working with

Read these before anything else.

1. What ComfyUI is and how it runs a graph
2. Nodes, inputs, outputs and links
3. Data types
4. Where ComfyUI files live
5. Workflow formats

#2. Your first picture

Then do the lesson A first picture from text and come back to these as the lesson names them.

1. Text to image
2. Checkpoint
3. Text encoders
4. Conditioning
5. Latent
6. Sampler and sampler names
7. Steps
8. Scheduler
9. VAE
10. Resolution and aspect ratio
11. Seed
12. Batch

#3. Steering with words

With the lesson The prompt: how a model reads it.

1. The prompt: how models read text
2. Prompt syntax and weights
3. The negative prompt
4. CFG and guidance

#4. Models and families

1. Model families
2. Why files of different families do not mix
3. Separate loaders
4. One page per family: SD 1.5, SDXL, Pony, Illustrious, SD 3 and 3.5, Flux.1, Krea 2, Flux.2, Chroma, Qwen-Image, Z-Image, Kolors, AuraFlow, Stable Cascade, Wan 2.1, Wan 2.2, HunyuanVideo, LTX-2.

#5. Picture techniques

With the lesson Refining with a second model.

1. Denoise
2. Image to image
3. Inpainting and masks
4. Outpainting
5. LoRA
6. ControlNet and preprocessors
7. Reference images
8. Upscaling

#6. Video and sound

With the lessons from Video from text on.

1. Video generation
2. Video length and frame rate
3. Wan 2.2: the two passes
4. Speed-up LoRAs
5. First and last frame
6. Video to video
7. Animation by pose
8. Talking portraits
9. Audio

#7. Working like a regular

1. Graphics card memory
2. Node packs
3. Unsafe model formats
4. Paid API nodes
5. Notes, reroutes and primitives
6. Bypass and mute
7. Subgraphs
8. Reading someone else's workflow

#See also

• ComfyUI concepts
• Lessons
• The first hour
• Glossary