Librum
Librum is a self-hosted e-books & digital libraries replacement for Evernote and Microsoft OneNote.
Cross-platform e-book library management, honestly reviewed. No marketing fluff, just what you get when you self-host it.
TL;DR
- What it is: Open-source (GPL-3.0) e-book reader and cloud library manager — think Kindle app, but the server runs on your hardware and Amazon can’t see your reading list [1][3].
- Who it’s for: Readers who want their e-book library synced across Windows, Linux, and macOS without surrendering their data to a closed ecosystem. Also privacy-minded users who’ve accumulated a large EPUB/PDF collection and want a clean, unified reader [3].
- Cost savings: Librum’s cloud plans run €1.49–€7.99/mo for the managed service. Self-hosted runs on any VPS or home server at infrastructure cost only — essentially zero for most setups that already have a server running [pricing page].
- Key strength: Genuine cross-platform sync with a consistent UI across Windows, Linux, and macOS, plus access to 70,000+ free public-domain books built in [1][3]. The only open-source option in this category that takes the multi-device reading flow seriously.
- Key weakness: It’s a native desktop application, not a web reader — you install it on each device rather than opening a browser tab. iOS and Android support is listed as planned but not yet shipped. Self-hosting requires running a separate server repo and editing config files or the Windows registry, which is meaningfully more friction than most self-hosted tools [README][3].
What is Librum
Librum is a native desktop e-book reader built in C++ that adds cloud sync, library management, and AI annotation tools on top of a conventional reading experience [2]. You install the client on each device, log in to your account (either their managed cloud or your self-hosted Librum-Server), and your library, bookmarks, highlights, and reading progress follow you across machines.
The pitch from the GitHub README is modest: “an application designed to make reading enjoyable and straightforward for everyone.” The more useful description is that it tries to give you what the Kindle app gives you — seamless cross-device sync, a unified library, a clean reader — without tying that experience to Amazon’s ecosystem or any proprietary cloud [3].
Where Librum differentiates: it bundles access to 70,000 free public-domain books (Project Gutenberg and similar sources) directly in the app [1][3]. It adds AI features for in-context explanation, summarization, and translation — gated behind the cloud subscription tiers [pricing page]. And it’s fully open-source under GPL-3.0, meaning anyone can audit the code or run their own server [README].
The project sits at 5,280 GitHub stars as of this writing, written primarily in C++, with a small core team and community contributors handling translations and bug reports [merged profile]. The website copyright reads 2024 and the project appears to be bootstrapped rather than VC-backed — relevant context if you’re evaluating long-term reliability [website].
Why people choose it
The GIGAZINE review [1] and the LinuxForDevices guide [3] tell the same story from slightly different angles: people come to Librum because they want a unified reading experience across operating systems without picking a vendor.
The cross-platform sync problem is real. LinuxForDevices names the exact frustration Librum is solving: “most [open-source e-book readers] are limited to the Linux operating system, so you will have to install a completely different application on a different OS such as Windows, macOS, or Android which would result in you dealing with inconsistent UI on different platforms” [3]. Calibre is the de facto standard for local library management, but it’s a desktop-only tool — your Calibre library doesn’t follow you from your Linux workstation to your Windows laptop without manual sync or third-party hacks.
The Kindle alternative angle. The implicit comparison is the Kindle ecosystem: Amazon’s app is seamless and cross-platform, but you don’t own the books, the app tracks your reading behavior, and your entire library can be revoked. Librum offers comparable sync convenience with a library you own and a server you control [3].
The free book catalog is a meaningful draw. The GIGAZINE walkthrough [1] spends significant time on this feature — browsing and downloading from 70,000+ free titles without leaving the app. For readers who primarily consume public-domain literature (classics, history, science), this isn’t a nice-to-have, it’s the whole app.
The AI annotation layer. The website pitches AI integration prominently: “Use it to explain, summarize or give you more information on any text within your book” [website]. In practice this is text selection → AI prompt, gated behind cloud subscription tiers (10 AI requests/day on free, 100/day on Pro at €2.99/mo) [pricing page]. It’s a real feature, not vaporware, but casual readers are unlikely to need more than the free tier’s daily limit.
What the sources don’t spend time on: any strong criticism of Librum itself. The foo.zone entry [4] is the most honest signal — a technically competent person listed Librum in their “self-hosting projects I don’t have time for” category. That’s not a knock on the software’s quality; it’s a realistic assessment that the self-hosted setup requires meaningful investment. More on that below.
Features
Based on the README, the GIGAZINE walkthrough [1], and the LinuxForDevices feature summary [3]:
Reader:
- EPUB, PDF, MOBI, and other major formats [2][3]
- Vertical scrolling layout with zoom controls [1]
- Full-text search within books [1][merged profile]
- Bookmarks with custom names, jumpable from a panel [1]
- Text highlighting with multiple colors and configurable transparency [3]
- Table of contents navigation [1][3]
- Full-screen / distraction-free mode [1]
- Text-to-speech support [3]
- Keyboard shortcuts throughout [3]
Library management:
- Account-based cloud sync — library, bookmarks, highlights, progress all follow you [1][3]
- Collections, tags, sorting, filtering by author and other metadata [README][3]
- Metadata editing for books in your library [3]
- Library view customization [3]
- 70,000+ free public-domain books accessible directly in the app [1][3]
- Personal reading statistics [README]
AI features (cloud-tier gated):
- In-reader AI for text explanation, summarization, additional context [website][pricing page]
- Translation feature (100–500 translations/day depending on tier) [pricing page]
- AI requests limited: 10/day free, 25/day Basic, 100/day Pro, 250/day Elite [pricing page]
Platform support:
- Windows, Linux, macOS native clients [README][1][3]
- Available on Flathub for Linux [3]
- iOS and Android listed as planned [README] — not yet shipped as of this writing
- Self-hostable via Librum-Server (separate repository) [README]
Plugin support listed as a canonical feature in the merged profile, though neither the README nor third-party articles elaborate on what the plugin API covers [merged profile].
Pricing: SaaS vs self-hosted math
Librum Cloud (managed service):
- Free: 1GB storage, 10 AI requests/day, 0 translations/day [pricing page]
- Basic (€1.49/mo): 5GB storage, 25 AI requests/day, 100 translations/day [pricing page]
- Pro (€2.99/mo, “Most Popular”): 20GB storage, 100 AI requests/day, 250 translations/day [pricing page]
- Elite (€7.99/mo): 150GB storage, 250 AI requests/day, 500 translations/day [pricing page]
Self-hosted:
- Librum-Server license: $0 (GPL-3.0) [README]
- Storage: whatever your server has — no per-GB fees
- AI requests: not applicable unless you wire in your own AI API (data not available on how this works self-hosted)
- Infrastructure: whatever you’re already paying for a VPS or home server
Kindle Unlimited for comparison:
- $9.99/mo for unlimited access to Amazon’s catalog — but you don’t own any of the books and the library disappears if you cancel
- Kindle app: free, but your purchases are locked to Amazon’s DRM and ecosystem
Concrete math for a typical reader: If you have 500 EPUB/PDF files in your personal collection and want them synced across three devices (workstation, laptop, reading-specific machine), the Pro tier at €2.99/mo (≈$3.20/mo at current rates) gives you 20GB — plenty for most text libraries. Over a year that’s roughly $38/year for the managed service.
Self-hosted on a server you already run costs you nothing incremental. Self-hosted on a dedicated €3.50/mo Hetzner VPS costs $42/year — basically the same as Pro, except you own the infrastructure and there are no request limits.
Where the math becomes significant: the Elite tier at €7.99/mo ($96/year) versus self-hosting. If you have a large library and heavy AI usage, self-hosting clears that gap quickly.
What’s missing from the comparison: there’s no clear documentation on whether self-hosted Librum-Server provides AI features at all, or whether those require the managed cloud. The README doesn’t address this, and none of the third-party articles tested it [README][1][3]. If you’re self-hosting primarily for AI annotation features, verify this before committing.
Deployment reality check
This is where Librum’s self-hosted story diverges from most tools in this category, and the difference matters.
Most self-hosted tools give you a web interface. You run Docker, point your browser at a port, done. Librum is a native desktop application. Self-hosting means running Librum-Server (a separate backend repo at github.com/Librum-Reader/Librum-Server) to handle sync, accounts, and storage — then configuring each client installation to point at your server instead of theirs [README].
The client configuration is not intuitive:
- Linux: Edit
~/.config/Librum-Reader/Librum.conf, setselfHosted=trueandserverHost=<your-url>[README] - Windows: Open the registry editor, navigate to
HKEY_CURRENT_USER\Software\Librum-Reader\Librum, and set the same two values [README]
The README notes you must run the application at least once first to generate the config files before you can edit them [README]. For a non-technical user, registry editing on Windows is a meaningful barrier.
What you actually need for the server:
- A server running the Librum-Server backend (Docker or bare metal — documentation is in the Librum-Server repo)
- HTTPS setup if you want to sync over the internet
- Network access from each client device to the server
The foo.zone signal. One technically literate blogger listed Librum in their “self-hosting projects I don’t have time for” alongside Pi-hole and their own Matrix server [4]. This is the most honest data point available on setup friction. It doesn’t mean the setup is broken — it means a competent sysadmin looked at the requirements and decided the benefit didn’t justify the time investment at that moment.
Platform reality: iOS and Android support is listed as future [README]. If your reading setup involves a tablet or phone as a primary device, Librum is not there yet. The Linux/Windows/macOS client support is real and working [1][3].
Realistic time estimate for a technical user: 1–3 hours to get the server running and clients configured. For a non-technical founder: hire someone or wait for a more turnkey install path.
Pros and Cons
Pros
- Consistent cross-platform UI. Windows, Linux, and macOS use the same client with the same feature set — a genuinely rare property in the open-source reader category [3]. The competition (Calibre, Foliate, Bookworm) is platform-specific or web-only.
- GPL-3.0 open source. The full source is auditable. No proprietary server components, no telemetry you can’t inspect [README][2].
- 70,000+ free books built in. Directly browse and download public-domain titles from inside the app. For readers of classics, this is a substantial library that requires no setup [1][3].
- Reasonable pricing tiers. €2.99/mo for the managed Pro tier is not extractive. The free tier is functional for light use [pricing page].
- Full reading feature set. Highlights, bookmarks, full-text search, TTS, TOC navigation, metadata editing — everything a serious reader expects [1][3].
- Active enough for current use. The GIGAZINE review (2024) and LinuxForDevices guide show a working, coherent product — not abandonware [1][3].
Cons
- Not a web reader. Every device requires a client installation and manual configuration to point at your self-hosted server. No browser fallback [README].
- Self-hosting is non-trivial. Two-step process: run Librum-Server, then manually edit config files or the Windows registry on each client device [README]. Technical users can do this; non-technical founders probably can’t without help.
- No iOS or Android. Mobile support is listed as planned but not shipped [README]. If your primary reading device is a phone or tablet, this is a blocker today.
- AI features may not work self-hosted. The pricing page gates AI requests behind cloud tiers. Whether self-hosted Librum-Server supports AI at all is undocumented across all reviewed sources [pricing page][README].
- GPL-3.0, not MIT. Stricter copyleft — you can’t embed Librum in a proprietary product or SaaS without open-sourcing your implementation. Not a problem for personal use, but relevant for developers [2].
- Small team, bootstrapped. The project has 5,280 GitHub stars and the website copyright sits at 2024 [merged profile][website]. Real, maintained software — but not a heavily-resourced project. Long-term continuity is a reasonable question mark.
- Pricing page discrepancy. The current pricing page shows 1GB on the free tier; a 2024 LinuxForDevices article says 2GB [3][pricing page]. Minor, but signals the documentation doesn’t always stay current.
- No native Calibre import path. If you’ve managed your library in Calibre for years, migrating metadata, collections, and reading progress to Librum is manual work. Data not available on tooling to assist.
Who should use this / who shouldn’t
Use Librum if:
- You have a large personal EPUB/PDF collection and want it synced across multiple desktops (Linux, Windows, macOS) with a consistent interface.
- You primarily read public-domain literature and want 70,000+ free books accessible without a separate Gutenberg workflow.
- You’re comfortable with Docker for the server side and don’t mind editing config files on each client once.
- You want to get off Kindle but don’t want to give up cross-device sync.
- You’re fine paying €2.99/mo for the managed Pro tier — the price is fair and removes the self-hosting friction entirely.
Skip it if:
- Your primary reading device is an iPhone, iPad, or Android phone. Mobile support isn’t there yet.
- You want a web-based reader you can access from any browser without installing anything. Librum is native-only.
- You’re non-technical and have no one to help with the Librum-Server setup. The client configuration step (registry editing on Windows, config file on Linux) will stop you.
- Your library is primarily audiobooks, comics, or manga. Librum is a text reader. Kavita and Komga serve those formats better [2].
- You need team or multi-user features — there’s no documented shared library or family account capability.
Stay on Kindle/your current setup if:
- You buy DRM-protected titles from Amazon, Apple, or Google and that catalog matters to you. Librum handles your own DRM-free files, not purchased DRM content.
- You want a product with dedicated mobile apps today, not eventually.
Alternatives worth considering
- Calibre + Calibre-Web — the established combination. Calibre manages your local library (conversion, metadata, organization); Calibre-Web puts a browser UI in front of it for remote access. More mature, larger community, web-based access means no per-device client install. More complex to set up than either component suggests. [2]
- Kavita — cross-platform e-book, manga, and comic server with OPDS support, web reader, user management, and ratings. GPL-3.0, .NET/Docker. Heavier feature set if you also read comics/manga; overkill for a text-only library. 132.7K estimated traffic versus Librum’s 27.5K suggests a significantly larger user base [2].
- Komga — media server focused on comics and manga with OPDS support and a web reader. Not optimized for EPUBs or PDFs [2].
- Readarr — book manager and automation tool. Handles acquisition and organization rather than reading. Often paired with Calibre-Web [2].
- Stump — fast, free, MIT-licensed comics/manga/digital book server. Rust-based, OPDS support. Smaller project (5K traffic), earlier stage [2].
- Standard Ebooks — not a reader, but worth mentioning: if you read public-domain literature, Standard Ebooks produces beautifully formatted, DRM-free EPUBs that work in any reader. Removes the need for Librum’s built-in catalog for many users.
For a non-technical founder managing a personal e-book library across desktops: the realistic shortlist is Librum versus Calibre-Web. Pick Librum if you want a polished native client and are willing to deal with the server setup. Pick Calibre-Web if you want browser-based access from any device and don’t mind Calibre’s dated but powerful desktop interface for library management.
Bottom line
Librum is the most serious attempt in the open-source space to replicate the Kindle app’s cross-device sync experience without Amazon. The reading feature set is complete, the UI is clean, and the pricing — whether managed cloud at €2.99/mo or self-hosted — is not extractive. The trade-offs are specific and real: no mobile apps yet, a self-hosting setup that requires more steps than most Docker-based tools, and a small team that may or may not sustain the project long-term. For readers with a desktop-primary setup who want their EPUB/PDF library synced across Linux, Windows, and macOS without signing anything with Amazon, it’s the right call. For anyone who reads primarily on a phone or tablet, come back in a year and check whether Android support shipped.
If the server setup is the blocker, that’s exactly what upready.dev deploys for clients — one-time fee, configured and running, you own the infrastructure.
Sources
- GIGAZINE — “Review of ‘Librum’, an e-book reader that allows you to read EPUB, PDF, and more than 70,000 books for free and sync them on Windows, Linux, macOS, etc.” (January 1, 2024). https://gigazine.net/gsc_news/en/20240101-librum-reader/
- Hosted Software — “Librum” — tool listing with alternatives and traffic data. https://hostedsoftware.org/tools/librum/
- LinuxForDevices — “Librum: The Cross-Platform Open-Source e-book Reader with Cloud Sync”. https://www.linuxfordevices.com/tutorials/linux/install-librum
- foo.zone — “Projects I currently don’t have time for” (May 3, 2024) — Librum listed under self-hosting projects deferred due to setup complexity. https://foo.zone/gemfeed/2024-05-03-projects-i-currently-dont-have-time-for.html
Primary sources:
- GitHub repository and README: https://github.com/librum-reader/librum (5,280 stars, GPL-3.0 license)
- Official website: https://librumreader.com
- Pricing page: https://librumreader.com/pricing
Features
Integrations & APIs
- Plugin / Extension System
Search & Discovery
- Bookmarks / Favorites
- Full-Text Search
Related Documents & Knowledge Base Tools
View all 226 →Stirling-PDF
75KThe most popular self-hosted PDF platform — merge, split, convert, OCR, sign, and process documents with AI, all running on your own infrastructure.
AppFlowy
69KAn open-source Notion alternative with AI, wikis, projects, and databases — cross-platform (desktop, mobile, web) with offline-first architecture and full data ownership.
AFFiNE Community Edition
66KAn open-source workspace that merges docs, whiteboards, and databases into one platform — a privacy-focused alternative to Notion and Miro with AI built in.
Docusaurus
64KA static site generator built on React for documentation websites — write in Markdown/MDX, version your docs, and deploy anywhere. Created by Meta.
Crawl4AI
62KOpen-source LLM-friendly web crawler that generates clean markdown from any website, purpose-built for RAG pipelines, AI data extraction, and automated research.
Atom
61KGitHub's hackable text editor, officially sunset in December 2022. The codebase remains archived on GitHub as a reference for community forks like Pulsar.