Is Librarium a self-hosted alternative to Goodreads?

Yes. Librarium runs on your own hardware (Docker or Kubernetes), keeps your reading data on your server, and has no ads, no parent-company tracking, and no vendor lock-in. The closest fit if you want a Goodreads or StoryGraph experience without sending everything to a cloud service.

How is Librarium different from Calibre or Readarr?

Calibre and Readarr are ebook-first — they exist to manage downloaded files. Librarium is print-first: it catalogs the physical book, manga, comic, or magazine on your shelf using a work/edition data model (FRBR-style), and treats ebook or audiobook files as optional references hanging off an edition. If your shelves are mostly physical, Librarium fits the way you already think about a collection.

Does Librarium have a mobile app?

Yes — a native SwiftUI iOS app, currently in TestFlight with the App Store release coming next. The iOS app includes barcode scanning so you can add books straight from a bookstore shelf. There is no Android client today, and no committed plan to build one.

Yes. Self-hosted Librarium is free and AGPL 3.0 licensed, and it ships today. Lite for iOS will also be free, and it is next up. There is no paid tier, and nothing is held back for one. You run it, you own the data. If you would rather not run a server yourself, a cloud host like Railway will run the same containers for a few dollars a month; that is between you and them, not us.

Does Librarium send my data anywhere?

Telemetry is opt-in and off by default. Self-hosted instances require an operator flag plus per-user consent before anything leaves the box, and the code that does the sending lives in the public repos. The only other outbound network calls are metadata lookups you actively trigger, like ISBN scans and cover backfills. You choose which providers to enable. Optional AI reading suggestions are off by default and use either your local Ollama instance or your own OpenAI or Anthropic API keys.

Can I import my Goodreads, StoryGraph, or Libib library?

CSV import from Goodreads, StoryGraph, and Libib is on deck; see the roadmap. Once shipped, you drop in the export from any of those services and Librarium will hydrate the catalog with metadata and covers from your configured providers.

Librarium — Self-hosted library tracker for books, manga & comics

Built for collectors who run their own stack.

Self-hosted

Runs on your infrastructure in Docker or Kubernetes. Telemetry is opt-in and off by default. No external data calls unless you ask for a metadata lookup.

Multi-library, multi-user

A single instance hosts independent libraries (shared household, personal collection, tech manuals) with per-library roles.

Print-first catalog

FRBR-style work/edition model tracks physical books, manga, comics, magazines, and technical documents without schema gymnastics.

Bring your own AI

Opt-in reading suggestions backed by Ollama, Osaurus, OpenAI, or Anthropic. Pick your provider and model. Run locally or ship it to the cloud, your call.

Scheduled work, observable

Cover backfills, metadata refreshes, and AI runs all live on a single cron-driven jobs page. Run on demand, inspect history, or let it tick quietly.

Bring your own storage

For ebooks and PDFs, Librarium stores a path reference. Your files stay where you put them.

Cross-platform

Go API, React web UI, and a native SwiftUI iOS app (TestFlight today, App Store soon).

Barcode scanning

Scan an ISBN at the bookstore from the iOS app. Instantly see if you already own it, or add it to any of your libraries.

Open source

AGPL 3.0. Ship patches, file issues, or fork it. Each client lives in its own repo and ships on its own cadence.

Editions

Two ways to run Librarium.

Both are free. Self-hosted ships today; Lite for iOS is next. Don't want to run a server? A cloud host will run the same containers for a few dollars a month, and deploy templates to make that a shorter job are on the way.

Lite

Coming soon

Free

Local storage with iCloud sync across your Apple devices, no server to run. The iOS app is on TestFlight today; Lite mode is shipping soon.

Works offline, syncs when online
Same data model as the server
ISBN lookup via Open Library
Point it at your own server any time

Try the iOS app

Self-hosted

Available

Free

Run it on your own server in Docker or Kubernetes. Your data stays on your network.

Full feature set, no caps
Multi-user with per-library roles
AI suggestions, MCP server
Opt-in telemetry, off by default

Deploy it

Compared

Pick the tool that fits your shelf.

A quick sanity check for readers weighing the options. An asterisk means the item is on the Librarium roadmap, not shipped yet.

Feature	Librarium	Calibre	Readarr	Bookwyrm
Self-hosted	✓	✓	✓	✓
Multi-user with per-library roles	✓	Basic	No	✓
Print-first catalog (physical books, manga, comics)	✓	Ebook-first	Ebook-first	✓
Native iOS app with barcode scanning	◐TestFlight	No	No	No
REST / OpenAPI	✓	Limited	✓	✓
AI suggestions (Ollama / Osaurus / OpenAI / Anthropic)	✓	No	No	No
Scheduled-jobs dashboard	✓	No	✓	No
Track books lent to friends and family	◐	No	No	No
Goodreads / Libib / StoryGraph CSV import	✓*	No	No	✓
Shared reading feed / activity	✓*household only	No	No	✓federated
Automated ebook acquisition	✓*	No	✓	No
Open source	✓AGPL 3.0	✓GPL	✓GPL	✓AGPL

Feature	Librarium	Goodreads	StoryGraph	Libib
Self-hosted — runs on your hardware	✓	No	No	No
Your data stays yours (no vendor lock-in)	✓	Amazonowner since 2013	Export	Export
Public API for integrations	✓REST · OpenAPI	Retiredkilled 2020	No	No
No ads, no data mining	✓	Ads	✓	✓
Works offline / LAN-only	✓	No	No	No
Print-first catalog (books, manga, comics)	✓	Books	Books	✓
Multi-user household with per-library roles	✓	No	No	No
AI suggestions under your control	✓your keys	No	No	No
Native mobile apps today	NoiOS only · TestFlight	✓	✓	✓
Barcode scanning	◐TestFlight	✓	✓	✓
Social reading feed (friends, reviews)	✓*household only	✓global	✓global	No
Goodreads / StoryGraph / Libib CSV import	✓*	No	✓from Goodreads	✓from Goodreads
Open source	✓AGPL 3.0	No	No	No
Cost	Free	Free	Free / $4.99Plus tier	Free / $9+paid tiers

◐ Partially shipped.* Planned roadmap item, not shipped yet. Comparisons reflect each project's upstream as of 26.4.

FAQ

Frequently asked questions.

Is Librarium a self-hosted alternative to Goodreads?: Yes. Librarium runs on your own hardware (Docker or Kubernetes), keeps your reading data on your server, and has no ads, no parent-company tracking, and no vendor lock-in. The closest fit if you want a Goodreads or StoryGraph experience without sending everything to a cloud service.
How is Librarium different from Calibre or Readarr?: Calibre and Readarr are ebook-first — they exist to manage downloaded files. Librarium is print-first: it catalogs the physical book, manga, comic, or magazine on your shelf using a work/edition data model (FRBR-style), and treats ebook or audiobook files as optional references hanging off an edition. If your shelves are mostly physical, Librarium fits the way you already think about a collection.
Does Librarium have a mobile app?: Yes — a native SwiftUI iOS app, currently in TestFlight with the App Store release coming next. The iOS app includes barcode scanning so you can add books straight from a bookstore shelf. There is no Android client today, and no committed plan to build one.
Is Librarium free?: Yes. Self-hosted Librarium is free and AGPL 3.0 licensed, and it ships today. Lite for iOS will also be free, and it is next up. There is no paid tier, and nothing is held back for one. You run it, you own the data. If you would rather not run a server yourself, a cloud host like Railway will run the same containers for a few dollars a month; that is between you and them, not us.
Does Librarium send my data anywhere?: Telemetry is opt-in and off by default. Self-hosted instances require an operator flag plus per-user consent before anything leaves the box, and the code that does the sending lives in the public repos. The only other outbound network calls are metadata lookups you actively trigger, like ISBN scans and cover backfills. You choose which providers to enable. Optional AI reading suggestions are off by default and use either your local Ollama instance or your own OpenAI or Anthropic API keys.
Can I import my Goodreads, StoryGraph, or Libib library?: CSV import from Goodreads, StoryGraph, and Libib is on deck; see the roadmap. Once shipped, you drop in the export from any of those services and Librarium will hydrate the catalog with metadata and covers from your configured providers.

Web

Browse, search, and curate from any browser.

The book catalog as a cover grid, with filters for ownership, library, read status and rating — Browse by cover
Narrow by shelf, library, list, read status or rating without leaving the grid.

The same catalog as a dense list, showing year, format and read status per row — Or by row
The same filters against a dense list, for when you want dates and formats at a glance.

Searching the whole instance from one box, with books and authors in the same result list — One search box
Books and the people who made them come back together, from anywhere in the app.

A single book with its cover, description, author, reading status and the lists it belongs to — The book itself
Description, contributors, reading status, and the lists it sits on.

One book expanded into its editions and physical copies, with ISBNs, format and condition — Work, edition, copy
A hardback and a paperback are two editions of one book, and every copy tracks its own condition.

Series cards showing how many volumes of each run are owned, and how many are missing — Series at a glance
Each run shows how much of it is on your shelf, and how much is not.

A single series with its volumes numbered in order, owned ones sharp and missing ones greyed out — Down to the volume
Volumes in order with the gaps greyed out, so you know what to hunt for.

Contributors across the shelves, each with a book count and covers, split by author, illustrator and translator — Everyone who worked on it
Authors, illustrators and translators are tracked apart, not flattened into one credit.

AI book recommendations split into what to read next and what to buy — Suggestions, if you want them
Off by default. Runs against your own Ollama instance or your own OpenAI or Anthropic key.

Adding a book by ISBN, showing the matched result and how many fields the provider filled — Add by ISBN
Providers fill what they can, and tell you how much of the record they filled.

Several books selected at once, with actions to set type, genre, tags, lists and refresh their metadata — Fix many at once
Select a range, then set type, genres, tags or lists, or send them all back for fresh metadata.

Metadata providers in priority order, each toggled on or off with its own API key — Your sources, your order
Providers are tried in priority order, and any of them can be switched off.

Background jobs on a cron schedule, with the next run time and a switch for each — Work that runs itself
Cover backfills, metadata refreshes and AI runs, on a cron you control.

Appearance settings with five themes, an OpenDyslexic typeface option and interface language — Five themes
Plus an OpenDyslexic typeface, and the interface in your own language.

iOS

Scan at the bookstore. One tap to see if you already own it, or add it to any library if you don't.

Roadmap

What's next.

Where each initiative sits in the pipeline. Solo project, so these are intentions not promises. Items move between columns as priorities shift.

Shipping next

Lite mode for iOS
iOS app redesign
Sync protocol (per-field timestamps, delta API)
App Store launch

On deck

CSV import (Goodreads, StoryGraph, Libib)
One-click cloud deploy templates
Bookstore links
Opt-in telemetry SDK
Series rework
Move books between libraries, in bulk
Higher-resolution covers

Exploring

Pre-ISBN and vintage collections
Member reviews on book pages
One list of every book across libraries
Edit metadata from the refresh panel
Shared reading feed (household scope)
iPad and Mac apps
AI metadata enrichment
Library invite links
Bulk select on books grid

Three repos, one product.

API-first: every client consumes the same OpenAPI contract. Each piece ships independently on a YY.M.R cadence.

librarium-api

Go backend · Postgres · River jobs

latest·

librarium-web

React + TypeScript + Tailwind + Vite

latest·

librarium-ios

SwiftUI · iOS 26+

latest·

Plus: chat with your library.

Optional companion service. Speak MCP to Claude Desktop, Cursor, Claude Code, or any MCP-aware client and let it search, read, add, and rate books on your behalf. Runs alongside your existing stack, authenticates with a regular API token, writes are gated by per-call approval in the client.

librarium-mcp

MCP server · Go · streamable HTTP

latest·