explaingit

karakeep-app/karakeep

📈 Trending★ 25,205TypeScriptAudience · vibe coderComplexity · 4/5ActiveLicenseSetup · hard

TLDR

Self-hosted bookmarking app with AI-powered tagging, full-text search, and page archiving for people who save lots of internet content and want to find it again.

Mindmap

mindmap
  root((Karakeep))
    What it does
      Save links and content
      AI auto-tagging
      Full-text search
      Page archiving
    How to access
      Browser extensions
      Mobile apps
      Web interface
    Key features
      Meilisearch integration
      RSS feed sync
      Video archiving
      Link rot protection
    Tech stack
      Next.js
      TypeScript
      Drizzle ORM
      tRPC
    Use cases
      Research organization
      Content curation
      Personal knowledge base
    Audience
      Self-hosters
      Privacy-conscious users

Things people build with this

USE CASE 1

Save articles, links, and PDFs from the web with one click and retrieve them later using AI-generated tags and full-text search.

USE CASE 2

Build a private personal knowledge base by archiving web pages locally and syncing content from RSS feeds and browser bookmarks.

USE CASE 3

Self-host a Pocket alternative on your own server to keep all your saved content private and under your control.

USE CASE 4

Automatically tag and organize research materials, news articles, and reference links without manual categorization.

Tech stack

TypeScriptNext.jsDrizzle ORMtRPCMeilisearchOpenAIOllama

Getting it running

Difficulty · hard Time to first run · 1day+

Requires Docker for Meilisearch, API key for OpenAI or local Ollama setup, and database initialization with Drizzle migrations.

Use it freely, but if you run it as a network service, you must release your changes to users. Strongest copyleft for SaaS.

In plain English

Karakeep (previously called Hoarder) is a self-hosted bookmarking app for people who save a lot of stuff from the internet, links, notes, images, and PDFs, and want to actually be able to find it later. The problem it solves is the familiar "I'll read this later" habit: you encounter interesting content on Reddit, Twitter, or news sites, save it somewhere, then can't remember where you put it or what it was about. What makes Karakeep stand out from a basic bookmark manager is AI-powered automatic tagging. When you save a link, it automatically reads the content and tags it with relevant categories using OpenAI or local AI models via Ollama. It also fetches the page title, description, and preview image so you can recognize items at a glance. A full-text search engine (powered by Meilisearch) lets you find content by searching the actual words in saved pages, not just titles. It archives full pages locally to protect against link rot (when websites disappear), and can auto-archive videos using yt-dlp. You get browser extensions for Chrome and Firefox for one-click saving, plus iOS and Android apps. It can also pull content automatically from RSS feeds and sync with browser bookmarks. You'd use this if you want a private, self-controlled alternative to tools like Pocket, especially if you self-host other apps and want your saved content to stay on your own server. Built with Next.js, TypeScript, Drizzle ORM, and tRPC.

Copy-paste prompts

Prompt 1
How do I set up Karakeep on my own server and connect it to my browser?
Prompt 2
Show me how to configure Karakeep to use a local Ollama model instead of OpenAI for tagging.
Prompt 3
How can I bulk import my existing bookmarks from Chrome or Firefox into Karakeep?
Prompt 4
What's the best way to search through my archived pages in Karakeep using full-text search?
Prompt 5
How do I set up automatic archiving of YouTube videos and other media in Karakeep?
Open on GitHub → Explain another repo

Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.