Overview / Description
Trelk — Read, Think, Connect is an AI personal knowledge management tool that saves, summarizes, and links everything you read directly on your device for researchers, students, and knowledge workers. Capture content by sharing URLs, scanning QR codes, importing PDFs and DOCX files, or writing notes, and let on-device AI automatically summarize, tag, and classify each item. Trelk combines full-text keyword search with semantic vector search, then maps your library as a force-directed knowledge graph that surfaces hidden connections and semantic clusters across articles, papers, and notes. You can chat with your entire library using Retrieval-Augmented Generation (RAG) with source citations, receive a daily AI briefing that synthesizes recent reading, and reinforce retention with AI-generated spaced-repetition flash cards. It works fully offline with zero configuration and can run on your favorite proprietary LLMs or open-source models from Hugging Face. Community collections let you import curated reading lists (via GitHub) or share your own. Trelk uses open formats and a buy-once, own-forever model with no subscriptions, accounts, or tracking — making it a private alternative for people who want a second brain that stays on their own hardware.
Used For
Researchers, students, and knowledge workers use Trelk to save, summarize, and connect their reading in a private on-device AI knowledge base.
Pricing
Plan
One-time purchase (buy once, own forever) with all features included and free updates
Pros & Cons
Pros
- On-device AI summarizes, tags, and classifies articles, PDFs, DOCX, and notes with no accounts or cloud tracking
- Hybrid search combines full-text keywords with semantic vector search across your whole library
- Force-directed knowledge graph visualizes hidden connections and semantic clusters between items
- Chat with your library via RAG with source citations, plus daily AI briefings and spaced-repetition flash cards
- Buy-once pricing with no subscriptions; works fully offline and supports proprietary or Hugging Face open-source models
Cons
- New product with a limited public track record and thin documentation
- Android/Google Play version listed as Coming Soon; currently focused on Apple's App Store
- Running local open-source models depends on device hardware, which may limit performance
- Niche appeal — geared to power readers and researchers rather than casual note-takers
Questions & Answers
Alternatives
Obsidian, Mem, Notion AI, Reflect, Readwise Reader, Heptabase
Reviews & Ratings
0 reviews
Sign in to rate and review Trelk — Read, Think, Connect.
Sign in to reviewNo reviews yet. Be the first to review Trelk — Read, Think, Connect!