18+ only. This article discusses adult content and is intended for adult audiences.
"Booru" sites — tag-based image boards descended from the original Danbooru — are the backbone of searchable adult art. Three names dominate: rule34.xxx, Gelbooru and e621. They look similar but serve noticeably different audiences. Here's how to pick.
rule34.xxx — the everything-library
The largest general-purpose adult booru, mixing anime, western cartoons, video games and comics. Its strength is coverage: if a character exists in any fandom, there are probably posts of them here. Tagging quality is decent but not strict, and the sheer volume means quality varies — which is exactly why score sorting and filters matter so much (see our guide to sorting by score). adultgpt is a search front-end for exactly this library, adding score sorting, blacklists, supertags and a faster interface on top of the same content.
- Example: the community's best Zelda posts — search on adultgpt
Gelbooru — the classic archive
One of the oldest and largest boorus, with a massive archive spanning both adult and safe content. Its API is widely cloned, which is why many "rule34 search" sites are Gelbooru front-ends in disguise. Coverage overlaps heavily with rule34.xxx, and the interface is famously dated. Use it if you want a second opinion on a niche tag that seems empty elsewhere — different uploaders mean different gaps.
e621 — the furry specialist with the best tags
e621 is the flagship booru of the furry fandom. Even if you're not its target audience, it's worth knowing: it has arguably the best tagging discipline of any booru — strict tagging guidelines, active tag wrangling and excellent disambiguation. Its tagging conventions (like extensive species and pose tags) have influenced every other site.
Which one when?
- Anime / game / cartoon character → rule34.xxx (or adultgpt, same library, better tools)
- Furry art → e621, no contest
- Niche western content or cross-checking → Gelbooru as a fallback archive
- Learning search itself → the tag systems are near-identical, so our syntax guide applies to all three
The common ground
All three share the same core model: five tag types, include/exclude/optional modifiers and community scores. Learn the system once and it transfers everywhere. The difference is the library behind the tags — pick the one that hosts your fandom, then let filters do the curating.