Self-Hosted / Projects / LinkWarden

LinkWarden
StableSpecs verified September 3, 2026 · against v2.16.2
Overview
Self-hosted bookmark manager with page archiving, reader mode, and collaboration.
LinkWarden is a self-hosted bookmark manager and read-it-later archive for people who want to save links, preserve the full page, and organize research without depending on a hosted bookmarking SaaS.
Why people choose LinkWarden
Official product pages frame LinkWarden around link rot, preservation, and the gap between a simple bookmark list and a true knowledge archive. It stands out because it can keep screenshots, PDFs, and HTML copies of saved pages while still feeling like a modern personal or team bookmarking tool.
What is included in the stack
Official docs and README describe collections and sub-collections, reader view, highlighting and annotations, collaboration and public sharing, RSS feeds, browser extensions, mobile apps, API keys, import and export, optional AI tagging, and optional Meilisearch-backed advanced search.
What to know before you deploy
The self-hosting docs are unusually direct: Linkwarden Cloud is the easiest starting point and may be cheaper than renting a VPS, but the self-hosted version still exposes the core feature set. For your own deployment, the official docs call out a comfortable baseline of 4 GB RAM, a 2 core machine, PostgreSQL for the standard Docker path, and Meilisearch if you want the richer search layer.
Best fit
Choose LinkWarden when you want a self-hosted Pocket or Raindrop style alternative that also preserves source pages for long-term reference. If you only need a minimal bookmark list with the lightest possible footprint, simpler tools may be easier to run.
Quick deploy
Paste into docker-compose.yml and rundocker compose up -d1services:2 postgres:3 image: postgres:16-alpine4 env_file: .env5 restart: always6 volumes:7 - ./pgdata:/var/lib/postgresql/data89 linkwarden:10 image: ghcr.io/linkwarden/linkwarden:latest11 env_file: .env12 environment:13 - DATABASE_URL=postgresql://postgres:${POSTGRES_PASSWORD}@postgres:5432/postgres14 restart: always15 ports:16 - '3000:3000'17 volumes:18 - ./data:/data/data19 depends_on:20 - postgres21 - meilisearch2223 meilisearch:24 image: getmeili/meilisearch:v1.13.325 restart: always26 env_file:27 - .env28 volumes:29 - ./meili_data:/meili_data
Configuration
Environment variables
| Variable | What it does | Example |
|---|---|---|
NEXTAUTH_URLrequired | Public auth callback base for the LinkWarden web app. | https://bookmarks.example.com/api/v1/auth |
NEXTAUTH_SECRETrequired | Required secret used by the authentication layer. | change-me-to-a-long-secret |
POSTGRES_PASSWORDrequired | Database password used by the official Docker Compose setup. | change-me |
MEILI_MASTER_KEY | Required when using the documented Meilisearch service for advanced search. | change-me-too |
Ports
3000/tcp— Default LinkWarden web interface port in the official Docker setup.
Volumes to back up
/var/lib/postgresql/data— Persistent PostgreSQL database storage for the official Compose setup./data/data— Primary LinkWarden data storage path used in the official Docker Compose file./meili_data— Persistent Meilisearch data directory when advanced search is enabled.
Interface previews
Screenshots of the LinkWarden interface.
Feature support
| Feature | Support |
|---|---|
| Full-page preservation in multiple formatsOfficial docs and README say LinkWarden stores saved pages as preserved copies, including screenshot, PDF, and HTML style archives. | Supported |
| Reader view with highlights and annotationsThe product site and docs describe a dedicated reading interface for saved content plus text highlighting and notes. | Supported |
| Collections, tags, and structured organizationOfficial materials show collections, sub-collections, tags, names, and descriptions for organizing saved links. | Supported |
| Sharing and collaborationThe platform supports public sharing and multi-user collaboration inside shared collections. | Supported |
| Browser extension and bookmark syncOfficial pages point to the browser extension and Floccus-based bookmark synchronization. | Supported |
| Advanced search and filtersSearch is core, but the self-hosting docs say advanced search needs a running Meilisearch service and related environment variables. | Partial support |
| RSS, API keys, and automation hooksOfficial docs mention RSS feed support and API access tokens for integrations and automation. | Supported |
| Optional AI taggingThe README calls out local AI tagging as an optional feature rather than a guaranteed default part of every self-hosted install. | Partial support |
Community signals
- Stars
- 20.1k
- Forks
- 863
- Open issues
- 680
- Last commit
- 1d ago
- Latest release
- v2.16.3
- Repo created
- Apr 2022
Includes open pull requests
1mo ago
Refreshed nightly from the GitHub API.
Questions
What is LinkWarden used for?
LinkWarden is used to save, tag, read, annotate, and preserve web pages so bookmarks do not disappear when the original content changes or goes offline.
Is LinkWarden a self-hosted Pocket or Raindrop alternative?
Yes. LinkWarden fits the same broad use case as read-it-later and bookmarking SaaS tools, but adds self-hosting and stronger page-preservation controls.
Does LinkWarden archive full pages?
Yes. Official docs and the README say LinkWarden saves preserved copies of webpages in multiple formats so you can still access the content later.
What hardware does LinkWarden need?
The official self-hosting setup guide says a 2 core machine and 4 GB RAM is a comfortable starting point, with about 1.5 GB for the image plus your data.
Does LinkWarden need PostgreSQL or Meilisearch?
For the standard Docker path, yes to PostgreSQL. Meilisearch is optional but required if you want the advanced search layer documented in the official setup guide.
Can LinkWarden share collections with other people?
Yes. The product and docs pages describe collaboration features, member permissions, and public sharing for saved collections and links.
What is the main LinkWarden tradeoff?
LinkWarden is richer than a basic bookmark tool, which is exactly why many people choose it. The tradeoff is that the fuller archive and search stack asks for more resources and more moving parts than a minimalist bookmarking service.
support // the lab
Found this write-up useful?
If it saved you time or a rebuild, you can support more practical homelab guides.