Learn almost anything, from scratch to proven mastery. AI gets better, so do you.
Download · Install · First steps · Changelog
Tell Terramentor what you want to learn and by when. The model you connect drafts the course and writes every lesson, diagram and question. Terramentor plans the course against your deadline, asks you to prove each topic rather than tick it off, and brings back what you are starting to forget. The app is free and has no account; the model is paid per use to its provider, under your own key, or runs on your own computer.
The pictures below come from one physics course for the Dutch pre-university exam (VWO).
Describe what you want to learn and the model drafts a course: phases, then topics, each with an overview, in your language. Add a book or your notes and the course follows their chapters, and its lessons draw on their pages. The draft runs in the background, so you can keep studying meanwhile. You can also import a course file or an Anki deck.
Give it a start date, a deadline and the days of the week you study. Bigger topics get more days, and pace is measured against that plan. Change the dates and whatever is still open is planned again.
The home page is a stream: lessons in short parts, a question after each part, flashcards, and questions from earlier topics you are starting to forget. Diagrams, charts and animations are written by the model as code and checked before you see them. The assistant, open beside any screen, teaches the topic in front of you, answers from your own files, looks at a photo or a document you attach, and can change a course for you, with Undo.
A topic is proven when you pass its mastery check (choices, numbers you work out, answers you type, steps you put in order), or when you work an exercise on paper, photograph it and have it marked. Before you see a question with one right answer, the model answers it again without the key, and a key it disputes is thrown out. That catches slips, not a misconception the model holds both times. You can always skip a topic; it then stays marked as skipped, not as known.
Flashcards come back just before you would forget them (scheduled with FSRS-6, fitted to your
own reviews once there are enough of them).
Anki decks come in with their media and review history and go back out as .apkg.
Every topic in your library on a planet you can turn, placed by what it is about rather than which course it belongs to. Pick a course and it replays the order you finished its topics in, and can save the replay as a video or a GIF.
A chatbot explains anything you ask, but keeps no plan, does not know what you have proved, and never asks you about last month's topic. Anki keeps what you put in it, but you write the cards. Terramentor puts the course, the deadline, the proof and the review in one loop.
Lessons and questions are written by the model you connect, pitched at what you told it about yourself. Switch to a newer model and the next lessons are better, with no app update.
A hosted model is the right choice for most people. Open models that fit an 8 GB graphics card write weaker lessons, slowly; the ones that teach well want 24 GB of video memory or more. Hosted is cheap here because a lesson is written once and kept: a fast, low-cost model such as GLM-5.3-Flash costs about US$0.20 per day of study.
The first screen connects one: Connect OpenRouter signs you in at openrouter.ai and brings back a key you own, or paste a key for any OpenAI-compatible service, or point it at Ollama, LM Studio or llama.cpp. Without a model the app still opens your courses, saved lessons and flashcards.
Your library is one SQLite file on your own disk, with a folder beside it for documents and card media. Nothing reports back to this project. What leaves the machine is what a lesson or an answer needs, sent to the model provider you chose, and, while a new course is drafted, web searches for reading on its topics (on by default, one switch in Settings). SECURITY.md lists every outbound connection.
Desktop. Download the zip for your system from the
releases page, unpack it and start
Terramentor.exe (Windows), Terramentor.app (macOS, Apple Silicon) or ./terramentor.sh
(Linux). There is nothing else to install. The launcher is unsigned, so Windows asks once
(More info → Run anyway) and macOS wants a right-click → Open the first time.
Docker:
curl -O https://raw.githubusercontent.com/ValeraZSD/terramentor/main/docker-compose.yml
docker compose up -dThen open http://127.0.0.1:3001.
From source (Node.js 22.19 or newer):
git clone https://github.com/ValeraZSD/terramentor.git
cd terramentor
npm install
npm run standaloneUpdates, the Docker password and phone access are in docs/INSTALL.md. The interface comes in twelve languages (docs/I18N.md).
- First steps: a five-minute tour.
- Install and update: desktop, Docker, from source, phone access, upgrades.
- Why it exists: what this is for, and what it is not.
- Architecture: React and TypeScript, Node and SQLite, and why. Start here to contribute.
- Technical reference: schema, engines, API.
- Security: threat model and every outbound connection.
- Contributing, Roadmap, Changelog.
Found a wrong lesson or a bug? Settings → General → About → Report a problem fills in GitHub's form for you to check and send.
Copyright © 2026 Valerii Kozhevets (ValeraZSD). AGPL-3.0-or-later, see LICENSE; third-party code is listed in THIRD_PARTY_NOTICES.md. The running app links to its own source under Settings → General → About.