Notes, PDFs, anything you read. We turn them into a knowledge base you can query — every answer cites its source.
Lose it, or grow it?
Save an article. Forget where. Search later. Nothing.
Pick a note from your vault. Ideas found, pages built, summary ready.
Link two notes by hand. One changes. The other stays old. The link dies.
Links stay alive. When new information contradicts the old, every related page stays current.
A year of notes. Just files. Every project starts over. Same reading. Same conclusions.
A year builds a wiki. Every new question draws on everything you've collected. Your knowledge grows.
One note. A world of connections.
Not features. A new relationship with everything you read.
Pick a note. The plugin reads, extracts, builds. No folders. No tags. No work.
Talk to your wiki like a friend. It reasons, then answers — naturally.
New files? Auto-ingested. Contradictions? Found. On your terms.
Aliases, duplicates, dead links, orphans, empty pages — all in one pass.
Deep dives on architecture, workflows, and model selection.
Qwen3.5, Qwen3.6, Gemma 4, oMLX — which model runs your wiki well, on what hardware, with what tradeoffs. A practical guide to going fully offline without sacrificing usability.
Pick a PDF from anywhere in your vault. The plugin reads it through your LLM provider, transcribes it to Markdown verbatim, and feeds the result into the standard wiki extraction pipeline.
Four phases that turn a natural-language question into a wiki-grounded answer. Why this pipeline is fundamentally different from vanilla RAG, where it borrows, and where it diverges.
Obsidian is all you need.
In Obsidian: Settings → Community plugins → Browse, search Karpathy LLM Wiki, click Install then Enable.
Open Plugin PageSettings → Karpathy LLM Wiki. Pick your LLM, enter key, test, save.
Cmd+P (or Ctrl+P). Type Ingest to add a single note or Multi-File Ingest to pick many at once. Type Query to ask.
Everything worth knowing, before you start.
You capture what you already see. The plugin reads every note, surfaces links you missed, and builds pages between them — across months, folders, and buried mentions.
No. The plugin only writes inside its own wiki/ folder. Mark a page reviewed and the AI never rewrites it — only adds. Every answer cites back to your notes.
Obsidian v1.11.0 or later (desktop) and an LLM provider API key. Long-context models work best. A single ingest runs $0.05 to $0.50 on the LLM side — the plugin itself is free and open source. Local models via Ollama or LM Studio need no key.
Yes. With Ollama or LM Studio, every byte stays on your machine. With cloud providers, data only leaves during the specific ingest or query you trigger — no telemetry, no backend, no storage.
PDFs are a first-class source — pick one from your vault and it joins the same wiki graph as your notes. You can also assign different models to ingest, lint, and query.
Open an issue on GitHub, or start a discussion. You can also leave a review on the plugin's Obsidian Community page.
Read the source. Verify every word.
Every line of the website and the plugin lives on GitHub. Fork it. Audit it. Make it yours.
The plugin writes only to its own wiki/ folder. Your notes are never read, edited, or replaced.
No tracking. No analytics. No server. The plugin never makes an outbound request on its own.
API keys live in your system keychain — never in plaintext, never in your notes.
Not an island. Every tool you love, now better.
See links
Wiki gives your graph meaning. Hub pages, clusters, orphans. Exploration, not decoration.
Save articles
Save the clip anywhere in your vault. Run `Cmd+P` → Ingest from folder. AI extracts, links, updates. One clip becomes 10+ wiki pages.
Query data
AI adds structure. Tags, dates, categories. Empty tables become living dashboards.
Track versions
Watch your knowledge evolve. Every commit tells the story of how you grew.
Make slides
Ask: "Summarize my research for a 10-minute talk." Wiki builds slides from your knowledge.
Visual canvas
AI builds concept maps, timelines, decision trees from your knowledge. No more blank canvas.
Twelve ways in — cloud, local, or your own endpoint. Switch any time.
Long-context models read your whole wiki in one pass — the complete picture, not fragmented snippets. Local models run fully on-device — no upload, no API key.
