# Danbooru tags

**Source:** https://promtable.com/glossary/danbooru-tags

> Danbooru tags are the structured tagging vocabulary inherited from the Danbooru imageboard — character, style, scene, expression tags — used heavily as the prompt language for anime-style diffusion models in 2026.

---
Danbooru tags are the structured tagging vocabulary inherited from the Danbooru imageboard — character, style, scene, expression tags — used heavily as the prompt language for anime-style diffusion models in 2026.

Anime-tuned diffusion models (NovelAI, Pony Diffusion, Animagine, AnythingV variants) are trained on captions in Danbooru-style tag format. Prompts in those models work best when they look like a Danbooru search query: comma-separated tags grouped by category (artist, character, copyright, general). The structure gives precise control — "1girl, blonde hair, ahoge, blue eyes, school uniform, masterpiece, best quality" is a typical anime prompt. Generic photoreal models (Flux, GPT-Image, Imagen) do not respond to Danbooru tags the same way; they expect natural language. Mixing styles in one prompt usually degrades both.

## When to use

- Prompts for anime-tuned diffusion models.
- Character / scene control via tag composition.

## Common mistakes

- Mixing Danbooru tags with photoreal-model natural-language prompts.
- Overloading tag count past ~80 — models start losing focus.

## Related terms

- [diffusion-model](https://promtable.com/glossary/diffusion-model)
- [negative-prompt](https://promtable.com/glossary/negative-prompt)
- [lora](https://promtable.com/glossary/lora)

*Last updated: 2026-06-01*
---

Original page: https://promtable.com/glossary/danbooru-tags
Maintained by Promtable (https://promtable.com). Content: CC BY 4.0. Cite as "Promtable — https://promtable.com/glossary/danbooru-tags".
Contact: info@vibecodingturkey.com.