Danbooru & Gelbooru Tag Merged | 142K Tags | Simplified Chinese

介绍

• 整合 Danbooru 与 Gelbooru 的标签数据库，截止日期2025-09，共计 142,571 个唯一标签。

• 适用于提示词整理、标签检索、LoRA/Dataset 标注辅助、标签翻译参考和本地工具开发。

文件说明

• tags_merged.csv

原始标签版，保留 tag、category、count 字段，alias 为空。

• tags_merged_zh.csv

简体中文翻译版，alias 列提供标签的中文翻译。

说明

• 数据来自 Danbooru 与 Gelbooru 的公开标签接口并按标签名合并。

• 中文翻译由本地翻译模型生成，专有名词、角色名和部分生僻标签可能存在不准确或保留原文的情况。

• count 小于 50 的标签已被过滤掉。

Introduction

• This dataset merges the tag databases from Danbooru and Gelbooru (current as of September 2025), comprising a total of 142,571 unique tags.

• It is suitable for prompt organization, tag retrieval, assisting with LoRA/dataset tagging, tag translation reference, and local tool development.

File Descriptions

• tags_merged.csv: Raw tag version; retains tag, category, and count fields; the alias field is empty.

• tags_merged_zh.csv: Simplified Chinese translated version; the alias column provides the Chinese translation for each tag.

Notes

• Data is sourced from the public tag APIs of Danbooru and Gelbooru and merged based on tag names.

• Chinese translations were generated using a local translation model; inaccuracies or instances where the original text is retained may occur with proper nouns, character names, and certain obscure tags.

• Tags with a count of less than 50 have been filtered out.