Roadmap

Where Mnemo stands

Mnemo is in active development and not feature complete. This page is the honest version: what you can rely on today, what the team is building now, and what is still a concept on the horizon.

Works today

Shipped in v0.6.5

Block based notes editor

Text, images, code blocks, Sketch diagram blocks, and PDF export.

Flashcards with spaced repetition

Multiple scheduling algorithms with keyboard driven review.

Mind maps

Styled nodes and edges, edit and preview modes, and a minimap.

Widget dashboard

Review stats, recent decks, and study goals on the Overview screen.

Global search and Quick Actions

Ctrl or Cmd K search, a command overlay, and editable keybinds.

Theming and localization

Light and dark themes, a theming system, and translation infrastructure.

In development

Work in progress

These are being built now. None of them are promises with dates attached. Some exist behind flags, some are partially built, and some are still early designs.

Mind maps overhaul

Reworking the mind maps system to make editing, navigation, layout, and studying smoother and more reliable.

Flashcards overhaul

Improving the flashcard system around card creation, review flow, scheduling clarity, and long-term maintainability.

Notes improvements

Continuing to improve the notes editor with better writing, editing, formatting, block handling (and types), and document stability.

Opening up the theme system

Making the theme system easier to extend so people can build and share their own Mnemo themes.

Widgets system

Building out the dashboard widget system so more useful study, planning, and progress widgets can live in one place.

Stability

Fixing crashes, smoothing rough edges, and improving reliability across the app.

Consistent UI and UX

Cleaning up interaction patterns, layout, visual consistency, and flows so the app feels more coherent end to end.

Local AI assistant

A local AI system built into the app and optimized for low-end hardware. Would power generative content, writing tools, teaching, and in-app assistance (tool calling and MCP). Built around orchestration and agentic AI.

Semantic search

Mind Search: an in-house cascading tree memory system for large PDF uploads and systematic access to user knowledge across notes, mind maps, and other parts of the app.

AI learning paths

Uses the local AI to turn uploaded files (.txt, .md, .pdf, .docx, .pptx, links, and more) or your own content (notes, mind maps) into chapters. Each chapter includes theory and practice modes, with optional exams before you move on.

Chat module

An interface and sidebar for accessing the local AI system.

Progress analytics

Locally stored learning data, combined with science and AI, to improve retention and surface insights. Opt-in enabled by default.

Future concepts

Not started yet

Ideas we are excited about but are not actively building yet. They may change significantly, merge with other work, or never ship.

Explore store

A future place inside the app for themes, extensions, and language packs. Every upload would be verified by us to meet safety standards.

Learning games

Developers building games and features from ecosystem data, such as flashcard-based games.

Audio and video

Audio Review would use offline AI to create engaging, realistic podcasts with proper SFX, music, transitions, and voices—designed for real learning. Local video through a custom creation engine and pipeline so AI can generate high-quality learning-style videos, with visual, sound, and transition libraries plus programmatic tools the AI can call (time sync and similar).

Latest release: v0.6.5

Highlights from the most recent release:

Text color formatting in notes, with a redesigned color picker
Image and Sketch blocks can be dragged directly to reorder them
New blocks can be inserted below any block type
Fixed a startup crash on Linux self contained builds
Faster spellcheck and smoother zoom in the notes editor

Full notes, known issues, and checksums are on the releases page. Recent editor work may still cause instability in some interactions, and we patch those as they are found. If you hit one, report it and it gets looked at.