Style Atlas: open-source ComfyUI nodes pack 1,600 styles and 827 characters into prompts
Dear-Spend-2865 · reddit · 2026-11-15
The author releases Style Atlas, three ComfyUI nodes that turn giant Excel sheets into structured prompts: Character Atlas (who), Setting Atlas (where), and Style Atlas (how it's rendered) — chain them, queue, done. Nodes default to bypass and fold down, keeping workflows tidy.
Scale of the data:
- 1,600 rendering styles (oil painting to Soviet storybook to "Acid Pulp Scratchwork") plus 120 modifiers for quality, genre, palette, and composition
- 827 known characters across 217 franchises with masked/unmasked and costume variants, plus 95 actor-in-a-role entries
- 185 facial expressions, 25 hand poses, 250+ body poses, 360+ outfits, 330+ accessories, 666 locations
Engineering highlights: seeded reproducible randomness everywhere, nested wildcards (drop adjective.txt into wildcards/ and write adjective), multi-subject style stacking, hover thumbnails, and everything in plain .xlsx — edit a row and restart ComfyUI. The author says Claude Opus 5.5 wrote ALL the code and helped clean thousands of descriptions; some vocabulary comes from Civitai community testing. Open-source on GitHub.
More from Multimodal
- mark_k: Killing Grok Imagine would be a huge mistake, ceding image/video to Chinese models — mark_k · 2026-11-15
- Omni Video Generation Space Trends on Hugging Face — CumaSystem · 2026-11-15
- AI glitches become a Seinfeld script: a hobbyist builds his own ComfyUI front-end — AxonkaiLab · 2026-11-15
- A 5,000-word Claude prompt wrangles 9 reference characters into a 20-second Subaru ad — Nimblecloud13 · 2026-11-15
- UChicago study: artists can engineer visual memorability, but computational models capture it better — plopesresearch · 2026-11-15
- Creator finds MiniMax H3 base R2V beats bf variants for voice cloning — Foreforks · 2026-11-15