0
Use ← → to navigate, C to copy name, ↓ to favorite, Esc to close
A visual reference for the Anima 2B anime image model: 43,220 Danbooru artist tags, each rendered once with the exact same prompt and settings, so you can see how the model interprets a given artist before you put the tag in your own prompt. The full generation recipe is on the Prompt DNA page, and an alphabetical index of all artists lists every tag with its Danbooru work count.
The same 43,220 artists were rendered twice. Anima Turbo 1.0 is the fast, few-step variant of the model; Anima Base 1.0 HD is the standard model at a higher-quality setting. Use the switch in the header to compare how each variant reacts to the same artist tag. Favorites and folders are shared between the two views.
Copy the name from a card and write it as @artist_name in your prompt, exactly as shown (spaces and escaped parentheses included). Sorting by Works favours artists with many Danbooru posts, which the model usually reproduces most reliably; Unique highlights styles that differ most from the average; Avg. Favs and Avg. Score follow Danbooru user ratings. The Jump box scrolls straight to artists with a given work count.
Where is the original site? The explorer was created by ThetaCursed; the original repository and its image CDN went offline in 2026. This is a maintained copy that adds the Turbo/Base switch and hosts the images itself.
Are my favorites stored online? No. Favorites and folders live in your browser (IndexedDB). Export them to JSON from the Favorites tab if you want a backup or to move them to another device.
Why are some previews similar? Artists with few Danbooru works leave a weak trace in the model, so their previews drift toward the model's default look. Higher work counts generally give a more distinctive result.
Which software does this work with? Any tool that runs Anima 2B: ComfyUI, Stable Diffusion WebUI forks and similar. The tag syntax is the model's own; the explorer only helps you pick tags.