Free, open-source Windows MVP

Your notes shouldn't just store knowledge, they should help you remember it.

A.R.K.A turns your Markdown and Obsidian notes into high-quality active-recall questions, then schedules reviews using spaced repetition—all locally, with no account required.

No account required • Local-first • Open source

Active Recall Knowledge Assistant

A.R.K.A Application Interface
The Problem

Your notes store knowledge. They don't help you retain it.

Knowledge Lost Without Review

≈90–92%

Knowledge fades without recall.

Reading and writing notes builds familiarity—not the ability to retrieve information later. Without active recall and scheduled review, most of what you write is forgotten.

→SM-2 schedules each question at the right time to reinforce long-term retention.

01
Question Creation Time

10–20 min → ≈1 min

Writing questions manually takes too long.

Identifying key concepts and writing effective questions for every note is repetitive and time-consuming work that most people skip entirely.

→A.R.K.A. generates questions from multiple notes in about one minute.

02
Graph-Guided Questions

Cross-Topic

Generic AI misses the connections.

Basic AI tools process each note in isolation, producing shallow fact-recall questions that miss how concepts relate across your knowledge base.

→A.R.K.A. maps concepts and relationships to generate deeper, test-worthy questions.

03
Why A.R.K.A. Is Different

Questions built from connections, not isolated text chunks.

A.R.K.A. maps concepts and relationships across your notes, then uses those connections to identify test-worthy knowledge and generate questions that assess real understanding—not just memorisation.

Finds the most test-worthy concepts
Connects related ideas across multiple notes
Generates factual, conceptual, and relational questions
Tests understanding beyond simple recall

Photosynthesis.md

Produces glucose and oxygen

Cellular Respiration.md

Uses glucose and oxygen

Knowledge Graph

producesproducesused byused byLight energyCO₂WaterChemical energyATPAerobic respirationChloroplastPhotosynthesisGlucoseOxygenCellular Respiration

Cross-Note Concept Question

“How do photosynthesis and cellular respiration complement each other in the transfer of energy and matter?”

Built from 2 notes · 4 connected concepts
Features

Everything you need to start remembering

Turn the notes you already have into active-recall questions and scheduled reviews.

Markdown & Obsidian Support

Import your existing Markdown notes and Obsidian vaults without changing your workflow.

AI-Generated Questions

Automatically generate high-quality multiple-choice questions from your notes for effective active recall.

Cross-topics Questions

Generate deeper questions by connecting related concepts across multiple notes and topics.

Local-First & Offline

Keep your knowledge private with offline access and optional cloud AI using your own API key.

Spaced Repetition

Reinforce long-term memory with intelligent review scheduling based on the SM-2 algorithm.

Recall Spaces

Organize approved questions into dedicated study collections for focused review sessions.

How It Works

Turn your knowledge base into a learning system

Step 01 — Choose Your AI
Step 01 — Choose Your AI

Build the Future of A.R.K.A With Us

A.R.K.A already generates questions from Markdown notes and schedules reviews. Next, we are expanding question formats, learning insights, file support, and cross-device access.

Built

  • Import Markdown folders and Obsidian vaults
  • AI-powered question generation from key concepts and relationships
  • Review, edit, regenerate, or delete generated questions
  • Organize approved questions into Recall Spaces
  • Offline review sessions powered by the SM-2 spaced repetition algorithm
  • Support local AI with Ollama or cloud providers using your own API key

Building

  • Additional question formats, including AI-scored open-ended, fill-in-the-blank questions
  • Learning progress, retention insights, and review history
  • Desktop review reminders and notifications
  • Continuous improvements to AI-generated question quality
  • Automatic desktop application updates
  • Adaptive assessment based on concept mastery and Bloom's Taxonomy

Exploring

  • Native support for macOS and Linux
  • Import from PDF, Word, plain text, and Notion
  • Secure cross-device synchronization
  • Support for additional learning materials, including slides and videos

Have a suggestion? Open an issue, contribute on GitHub, or reach us at hello@akra.app