Illustrious & NoobAI Style Explorer: Now with 16,000+ Danbooru Artist Aesthetics (Free, Open Source, Online/Offline)
Free, open-source tool now maps 16,000+ styles for AI image generation.
ThetaCursed has shipped the final update of the Illustrious & NoobAI Style Explorer, a free, open-source visual reference library for Danbooru artist tags. The update triples the database to over 16,000 styles by adding 11,000+ new aesthetics, mapping the full stylistic potential of the Illustrious XL and NoobAI-XL models. The tool is designed to show the "pure DNA" of an artist's style using a neutral baseline generated with Nova Anime XL and no quality tags like "masterpiece" or year modifiers. It also recalculates uniqueness scores to surface the most distinct and expressive styles at the top, saving users from endless scrolling. The project includes a master list of 33,000 compatible artist tags filtered by training cutoff dates for power users.
The tool works instantly on desktop and mobile browsers, with a lightweight interface that supports 1-click tag copying, full-screen swipe browsing with hotkeys, and offline use via a ~900MB download. Users can sort favorites into custom folders and export them as .txt or .json. ThetaCursed notes that testing lower post-count tags revealed a clear limit—for every 3 recognizable gems, there are roughly 7 "empty" styles that the models do not distinctly recognize. Further expansion would dilute quality, making this the final update. The full project is available on GitHub, and the original thread on Reddit's StableDiffusion community has garnered significant attention.
- 16,000+ total styles: Tripled the database with 11,000+ new artist aesthetics for Illustrious XL and NoobAI-XL
- Recalculated uniqueness scores: Most distinct styles are now easier to find at the top of the list
- Fully offline and free: ~900MB download runs locally via any desktop browser with 1-click tag copying
Why It Matters
Maps the full stylistic potential of two major AI models, giving creators a complete, high-performance reference library.