Quick character concept preview using 12-step generation. Use when the user wants to rapidly iterate on character designs -- explore different looks, outfits, poses before committing to full quality.
Quick 12-step character concept preview optimized for character design iteration.
comfyui-generate-image or comfyui-portraitcomfyui-portraitcomfyui-img2img-remixcomfyui-preview-imagepython3 scripts/run.py "<character description>" <output-path> [options]
| Option | Default | Description |
|---|---|---|
--width | 512 | Image width |
--height | 768 | Image height (vertical for characters) |
--steps | 12 | Sampling steps (kept low for speed) |
--cfg | 5.0 | Classifier-free guidance scale |
--seed | random | Seed for reproducibility |
--negative | "watermark, text, blurry, low quality, deformed, extra fingers, multiple people, crowd" | Negative prompt |
# Quick character concept
python3 scripts/run.py "female elf ranger, leather armor, bow, forest background" elf_concept.png
# Iterate on outfit variations
python3 scripts/run.py "cyberpunk hacker, neon jacket, goggles, dark alley" cyber_v1.png --seed 42
python3 scripts/run.py "cyberpunk hacker, trench coat, visor, dark alley" cyber_v2.png --seed 42
# Fantasy warrior concept
python3 scripts/run.py "dwarf warrior, heavy plate armor, battle axe, mountain backdrop" dwarf.png
Output can feed into:
comfyui-generate-image -- once design is finalized, generate at full qualitycomfyui-portrait -- generate a high-quality portrait versioncomfyui-download-image -- download the preview"I'll generate a quick character preview of '[description]' at 512x768 (12 steps, ~10 seconds). Go ahead?"
| URL | Purpose | Data sent |
|---|---|---|
$COMFY_URL/prompt | Submit workflow | Text prompt, dimensions, sampling params |
$COMFY_URL/history/<id> | Poll job status | Prompt ID |
$COMFY_URL/view?filename=<f> | Download image | Filename |