Drag and drop an image, or click to upload
Or copy an image and press Ctrl+V in this area
Extracted Tags
This is not extracted from your image — edit it however you need.
AI image tagging · NAI・Danbooru-style image tag extractor
Drag and drop an image, or click to upload
Or copy an image and press Ctrl+V in this area
This is not extracted from your image — edit it however you need.
AnimeTag-AI is a free image tag extractor built for AI image tagging. Upload an illustration generated with NAI (NovelAI) or any subculture/anime-style artwork, and an image tagger AI model trained on Danbooru image data will extract General tags covering character appearance (hair color, eye color), outfit, pose, and art style, along with character/series tags and a content rating — all in a single pass. It's designed to be useful for AI art creators, NAI users, and anyone preparing a dataset for LoRA training.
No sign-up or installation required — it runs right in your browser. Uploaded images are processed only temporarily for tag inference and are never stored on our servers.
Automatically extracts standard Danbooru tags like 1girl, blue_hair, and school_uniform, each with a confidence score.
Character and series names are shown in their own separate category, cutting down on dataset labeling time.
Automatically classifies images as General/Sensitive/Questionable/Explicit, so you can safely filter your dataset for its intended purpose.
Adjust sensitivity with the slider and the tag list updates instantly — no server round-trip needed.
Copy tags as a ready-to-use, comma-separated prompt for Stable Diffusion or LoRA caption files.
Just drag and drop or paste from your clipboard to get started. No installation, no payment.
Drag and drop your illustration, or paste an image copied to your clipboard with Ctrl+V.
Use the slider to adjust detection sensitivity and filter the tag list to just what you need.
Click any unwanted tag to remove it, then use the Copy Prompt button to grab your results.
It's a free web service that uses an image-to-tag AI model to automatically extract Danbooru-style tags and a content rating from any subculture or anime-style illustration you upload.
We use an AI model trained for the image-to-tag task. It was trained on Danbooru image data, so it's specialized for recognizing subculture illustrations.
Yes. NAI and most other AI art generation tools are trained on the Danbooru tag system, so this service can extract image tags from them just as well.
No. Uploaded images are only processed temporarily for tag inference and are not separately stored on the server. See our Privacy Policy for more details.
Yes. Extracted tags can be copied directly as a comma-separated prompt, ready to use in Stable Diffusion LoRA caption files.
Yes. You can use it for free, with no sign-up and no payment required.