Is this an AI for Logseq that works offline?

Yes. Point a vault at your graph folder: the Markdown files are read and indexed on your machine. With a local model, downloaded once, the answers are written offline too. A cloud model is a choice you make per conversation.

Do I need a plugin in Logseq?

No plugin exists, and none is needed. Mnemosyne OS watches the folder where Logseq keeps your files and picks up every save. Logseq does not know it is there.

Does it write into my graph?

Never. It reads your notes and keeps its index in its own vault, alongside. Your files do not move a byte.

Which files does it read?

Markdown (.md, .mdx), plain text, reStructuredText and LaTeX, plus the PDFs, Word files and spreadsheets that sit in the same folder. Images once image memory is on.

Can I keep some notes out of the AI's answers?

Yes. Put them in their own vault and mark it MAXIMUM: it answers only questions that name it, and its content is not sent to a cloud model unless you confirm it, each time. AES-256 encryption at rest exists too; you arm it.

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 local model. Cloud credits are prepaid and optional.

For the Logseq community

Stay on Logseq.
Give your graph a memory.

Mnemosyne doesn't replace your graph, it plugs into it. Your markdown blocks and pages, right where they live, become a memory you query in one sentence: by meaning, across years of notes, and 100% local. (Logseq, Obsidian, or a plain folder of .md, if your notes are files, it reads them.)

Download, free How it works

We speak the same language

Local-first, plain files, zero lock-in.

That's your DNA. It's ours too. Your notes are markdown files on your disk, Mnemosyne reads them right where they are, no account, no server, no mandatory cloud. You can unplug everything tomorrow: your files haven't moved a single byte.

No proprietary lock, no export to negotiate. The same standard that made you pick files over a closed database.

What changes

From "where did I already write this?" to "just ask".

Keyword search finds what you already remember. Retrieval by meaning finds what you forgot you wrote, the note from two years ago that answers today's question, even without the right word.

And this memory doesn't sit in a corner: it feeds the AI you plug in, yours, local, or via your own key. Your assistant stops starting from scratch. It works from what you've already thought.

What it looks like

Your notes, as living memory.

Your notes, rendered as a memory you can walk through: each node a note, each edge a measured link of meaning. What you wrote no longer sleeps in a folder, it recalls itself.

Questions

What Logseq users ask before installing.

Is this an AI for Logseq that works offline?: Yes. Point a vault at your graph folder: the Markdown files are read and indexed on your machine. With a local model, downloaded once, the answers are written offline too. A cloud model is a choice you make per conversation.
Do I need a plugin in Logseq?: No plugin exists, and none is needed. Mnemosyne OS watches the folder where Logseq keeps your files and picks up every save. Logseq does not know it is there.
Does it write into my graph?: Never. It reads your notes and keeps its index in its own vault, alongside. Your files do not move a byte.
Which files does it read?: Markdown (.md, .mdx), plain text, reStructuredText and LaTeX, plus the PDFs, Word files and spreadsheets that sit in the same folder. Images once image memory is on.
Can I keep some notes out of the AI's answers?: Yes. Put them in their own vault and mark it MAXIMUM: it answers only questions that name it, and its content is not sent to a cloud model unless you confirm it, each time. AES-256 encryption at rest exists too; you arm it.
Is it free?: 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 local model. Cloud credits are prepaid and optional.

Your graph already knows everything. It just couldn't remember it.

Download, free Why the relationship