Mnemosyne OS
New The documentation is live: every engine, step by step. docs.mnemosyne-os.io →

Students

Your whole course, searchable the same evening.

The lecturer's PDF. Your photos of the projector. The notes you typed in Word. One memory, on your own computer. You ask a question, and the answer cites the slide, the lecturer and your note.

The problem

Three traces of one lecture, and nothing that ties them together.

The PDF arrives before or after. The photos sit on your phone. The notes are in a Word file named after the date. At exam time you have three piles and no thread between them.

Mnemosyne OS reads all three into one memory. Not a summary written for you. Your own material, findable by meaning and by the exact words.

The gesture

Copy the first words of the slide. That is the whole method.

When a slide comes up, start your note with its first words, exactly as written. Then write what the lecturer says. The PDF contains those words. The photo too, once read. Your note now has an address.

Ask about that point later, and the three passages come up together: what the slide said, what the lecturer added, what you wrote down.

photo of slide 12 Photosynthesis in three stages
lecture-notes.docx Photosynthesis in three stages he insisted on the light phase, exam question
chapter-3.pdf, page 12 Photosynthesis in three stages
The same first words in the photo, the note and the PDF. Recall brings the three together.

In the evening

Ask the way you would ask a classmate who wrote everything down.

  • What did the lecturer say about the light phase?
  • What exactly did slide 12 say?
  • Where do my notes and the PDF disagree?
  • Write me a revision sheet for chapter 3 from my notes and the PDF.

Free to study

No account. No subscription. The complete OS, for free.

You download, you launch. Nobody asks your name or your email. Everything on this page runs without a license and without credits, with a model on your own machine.

Your courses stay on your computer. With a local model, nothing leaves. With a cloud model, only the passages that answer your question are sent, and you choose that.

How it works

Five steps, once. Then it just runs.

  1. 01 One vault, one folder "+ New Vault", name it Courses, then "Add a folder to watch…". One subfolder per subject.
  2. 02 Turn on image memory Settings → Images. Add "Read their text" if you want the words on your photos.
  3. 03 In the lecture hall The slide's first words at the head of your note. Word or the built-in Notes.
  4. 04 After the lecture Drop the PDF and the photos in the folder. They enter on their own.
  5. 05 In the evening Set the scope to Courses and ask. The answer cites your files.

The detailed walkthrough lives in the documentation: Your first vault ↗ ·Image memory ↗

Questions

What students ask before installing.

How do I analyse my lecture notes with an AI?
Put your notes in a folder watched by a Courses vault. They enter memory at every save. Then ask in the chat: what the lecturer said about a point, what your notes miss compared to the PDF, a revision sheet for a chapter. The answer draws on your files and names them.
Does it read my Word files?
Yes. Save the .docx in the subject folder and its text enters memory at every save. Word's temporary files are ignored. Pictures pasted in the document are recovered if you turn on "Extract the images inside Word documents" in Settings → Images.
What about my photos of the projector?
Yes, once image memory is on in Settings → Images. The text written on the photo is read if you install a free OCR engine and click "Read their text". Formats: jpg, png, webp. The iPhone HEIC format is not read: set the camera to its most compatible format.
Why copy the first words of the slide?
Because memory searches by meaning and by the exact words. The lecturer's PDF contains those words, your note does, the photo does once read. The three passages then come up together. Without it, "they stressed the second point" points at no slide at all.
Is it free for students?
The download is the complete OS, free, with no account and no email. Everything on this page runs without a license and without credits, with a model on your own machine. If you prefer a cloud model, you load prepaid credits of 10, 25 or 50 $, or use your own API key.
Do I need a powerful computer?
It runs from 8 GB of RAM, in a lighter mode with a cloud model. Everything local is comfortable from 16 GB. The download page details the configuration. System requirements →
Do my courses stay private?
They stay on your computer, in files you can read without us. With a local model, nothing leaves. With a cloud model, only the passages that answer your question are sent, to the provider you chose. There is no account, so there is no database to find you in.
I take my notes in Obsidian. Does it work?
Yes: an Obsidian vault is a folder of Markdown files, and you give it to a vault to watch. Logseq, Notion and the others are covered on the page about connecting your notes. Connect your notes →