Media & libraries
Three dedicated media types — books, CDs, DVDs/Blu-rays — each with its own purpose-built fields instead of generic free text. Any number of independent libraries, shared per user with fine-grained control.
Early beta · actively developed · AGPL-3.0 licensed
A library for books, CDs and films — run by you, hosted on your own server.
MedInv catalogs physical collections automatically: scan a barcode, let it fetch title & cover art through metadata lookups, organize everything across any number of libraries — and search, analyze, and share it with others whenever you like. Runs entirely on your own infrastructure.
Feature set
Everything a serious collection needs.
Six areas, one tool — from the first scan to the last backup.
Capture & metadata
Scan a barcode or type a title — MedInv queries up to nine providers per media type automatically and downloads covers locally. A miss is never a dead end: manual entry always stays an option.
EAN/ISBN scan
Optional AI lookup
Cover download
Search & statistics
Full-text search with genuine typo tolerance, fine filters by genre, language, year, or price. Plus distributions and value history per library — including automatic currency conversion.
Access & sharing
Three account levels — guest, user, admin — plus fine-grained sharing per library. Optional OpenID Connect/OAuth2 login alongside classic accounts, with brute-force protection.
Backup, export & import
Automatic and manual backups with configurable retention, plus a safety backup before every schema-changing update. Full instance restore or targeted per-library export.
Languages & themes
18 bundled language packs, six themes, and its own CSS plugin system for a custom look — switchable per instance, with an editor for adding further languages.
In practice
This is what your own collection feels like.
Real interface, real sample data — no mockups.
Search
Find any title again, no matter how large the collection grows.
A full-text search across every library you can see, combinable with structured filters — plus a genuine, typo-tolerant fuzzy search for when the name isn't quite on the tip of your tongue.
- Filters by media type, genre, language, medium, price, year
- Saved searches for recurring queries
- Results automatically respect library sharing
Capture
Scan instead of type.
A single barcode is enough: MedInv queries every enabled metadata provider in parallel and merges the results field by field. No matching code on hand? Free-text title search works just as well.
- Camera scan right in the browser, or manual entry
- Bulk import of whole lists from a text file
- Re-fetch metadata for any title later
Libraries
Clear lists instead of index-card chaos.
Every library shows its holdings as a sortable, exportable list — including PDF export, multi-select for bulk changes, and direct access to every detail.
- Bulk edit and delete for selected titles
- PDF export of an entire library
- Free-text shelf location per copy
Plugins
Every metadata provider, controlled individually.
More than twenty metadata providers across all media types — API-based, web-scraping, or AI-powered — can each be enabled, configured, and tested individually, with no code changes required.
- Its own API key & settings per provider
- EAN support per provider visible at a glance
- Connection test right in the interface
Reports
Answers to concrete questions about the collection.
Seven ready-made tables answer what a chart alone can't: where are the duplicates? Which entries are missing a cover or price? What was added most recently?
- Duplicates, data quality, top lists, sharing & more
- Every table exports to PDF
- Spans all libraries, respects sharing
Statistics
The collection in numbers, sorted by frequency.
Every library gets its own distributions — genre, language, director, medium, release year — plus a value history over time and automatic currency conversion for mixed holdings.
- Distributions always sorted by frequency
- Value history over time, per library
- Automatic conversion across multiple currencies
Quick start
One container, no setup marathon.
MedInv runs as a single, self-contained Docker image — with an embedded SQLite database, no separate database container required.
Terminal
# Then available at http://localhost:8080
docker run -d \
--name medinv \
-p 8080:8080 \
-e MEDINV_ADMINUSER=admin@example.com \
-e MEDINV_ADMINPASS='ChangeMe123!' \
-v medinv-storage:/var/www/backend/storage \
--restart unless-stopped \
ghcr.io/vulture20/medinv:latest
The medinv-storage volume is what makes the database, covers, and backups survive restarts — always mount it.
Every option, including MariaDB/PostgreSQL support, is documented in the
README on GitHub.
Backend · PHP / Laravel 13
Frontend · React / TypeScript
Database · SQLite, MariaDB, or PostgreSQL
Deployment · one Docker image, one port