Danbooru 标签超市

Danbooru 标签超市 is a web-based reference and assembly tool AImed at people who write Danbooru-style Prompts for anime Image Generation models. Instead of mEMOrizing tag vocabulary, users can browse a curated catalog of tags, see which ones come with example illustrations, and combine them into prompt strings before pasting them into their preferred generation interface.

What it does

The site organizes a large tag library into a searchable collection. At the time of its documentation, it listed 3135 tags, of which 1444 included accompanying images, so users can visually confirm what a tag ACTually represents rather than guessing from the name alone. It also bundles preset tag groups, embedding model entries, and hypernetwork model entries that pair with common anime checkpoints.

Key capabilities

• Illustrated tag browsing — compare tags against reference images to reduce prompt guesswork and mismatched ouTPUts.
• Preset tag groups — 21 prebuilt groups let users quickly drop in common quality, style, or composition modifiers.
• Embedding model list — 22 embedding models are cataloged for reference when selecting textual inveRSIon assets.
• Hypernetwork list — 2 hypernetwork models are listed for users who work with that older fine-tuning fORMat.
• Prompt assembly — tags can be gathered and combined into a ready-to-use prompt string.

Who it is for

This tool suits anime-focused image generation hobbyists, prompt engineers, and artists using Stable Diffusion-style interfaces who want faster, more accurate tag selection. It is most useful for people already familiar with the Danbooru tagging convention, since the catalog assumes that vocabulary. Beginners may find it helpful as a visual glossary, but should expect a learning curve around how tag order and weighting affect results.

How to get started

1. Open the site in a desktop browser.
2. Search for a concept, character type, or style and review the illustrated matches.
3. Add chosen tags to your working set, optionally starting from one of the preset groups.
4. Review the embedding and hypernetwork entries only if your Workflow uses those formats.
5. Copy the assembled tags into your image generation tool and refine from there.

Pricing and access caveat

The site APPears to be freely accessible through a web browser with no account required, though this may change. The interface is presented primarily in Simplified Chinese, and English support is not guaranteed. Tag counts and model lists reflect a snapshot and may vary as the catalog is updated. Because the site functions as a reference layer rather than a generator, output quality still depends entirely on the model, checkpoint, and settings used elsewhere.