A self-hosted app for searching anime torrents by title or any major database ID.

Requires Docker.

docker volume create kbdex-data

docker run -d \
  --name kbdex \
  -p 8000:8000 \
  -v kbdex-data:/app/data \
  --restart unless-stopped \
  ghcr.io/kingbenny101/kbdex:latest

Or with Docker Compose:

git clone https://github.com/kingbenny101/kbdex.git
cd kbdex
docker compose up -d

The app will be available at http://localhost:8000. API docs are at /docs.

Requires uv.

uv sync          # install dependencies from uv.lock
uv run pytest    # run tests
uv run kbdex     # start the server on 0.0.0.0:8000

Or use the Makefile targets: make dev (auto-reload), make test, make up.

Search by title or any anime database ID — AniDB, MAL, AniList, Kitsu, TVDB, and more. kbdex resolves it to an AniDB ID, looks up all known title variants, searches the configured indexers, and parses each torrent filename to filter and rank results by season and episode.

MAL ID / AniList ID / ...          free-text query
        │                                 │
        ▼                                 │
 anime-lists mapping                      │
        │                                 │
        ▼                                 │
    AniDB ID  ──►  resolve titles         │
                        │                 │
                        └────────┬────────┘
                                 ▼
                          search indexers
                                 │
                                 ▼
                     parse filenames (anitopy + guessit)
                                 │
                                 ▼
                       filter by season / episode
                                 │
                                 ▼
                           JSON results

At startup, kbdex downloads two data files and refreshes them weekly:

• AniDB titles dump — maps AniDB IDs to all known title variants (romanised, Japanese, English, synonyms)
• anime-lists (Fribb/anime-lists) — maps MAL, AniList, Kitsu, TVDB, AniSearch, ANN, LiveChart, and Simkl IDs to AniDB IDs

At search time, the resolved titles are sent to each indexer. Every torrent filename is parsed with anitopy (with guessit as a fallback) to extract episode number, season, resolution, codec, and release group. Results are filtered to match the requested season/episode and sorted so exact episode matches rank above batch releases.

All settings are stored as XML files under the data volume and can be edited directly or via the web UI at /ui/settings.

Config files are created with defaults on first start. There are no environment variables.

Both are enabled by default. Each can be excluded per-request via the indexers parameter, or disabled entirely by setting an unreachable base URL in its config file.

This tool returns metadata from public torrent indexers. No content is hosted or proxied. Use responsibly and in accordance with the laws of your jurisdiction.

MIT