Updates, guides, and deep dives about the plugin and the Karpathy LLM Wiki concept.
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.
A deep dive into how Karpathy LLM Wiki's v1.23 graph engine uses Monte Carlo Personalized PageRank to walk your [[wiki-link]] graph in constant time per query — and why a 2,000-page vault costs the same as a 200-page one.
Three files change. The Karpathy LLM Wiki plugin you use doesn't. Why the project is migrating to Apache 2.0 with DCO — and what it means for users, contributors, and forks.
Karpathy LLM Wiki walks your wiki's [[link]] graph via Monte Carlo Personalized PageRank to find what you wrote about. Plus streaming for every LLM provider, multi-file ingest, and an auto-generated Welcome note.
Web Clipper saves the entire AI session to a folder you choose; Karpathy LLM Wiki extracts entities and links them across platforms into one Obsidian knowledge base.
From citation manager to queryable research knowledge base. Karpathy LLM Wiki sits at the end of a Zotero-to-Obsidian automation chain for academic literature.
Canvas reveals structure the Graph View can't — clusters, hubs, and orphans. Here's how Karpathy LLM Wiki flows into Canvas so your wiki becomes a spatial map instead of a node cloud.
How to turn academic PDFs into a structured research knowledge base in Obsidian. Cross-referenced entities, concepts, and theories that Karpathy LLM Wiki makes searchable.
Clip any web article into Obsidian, then watch Karpathy LLM Wiki turn it into a connected wiki page with bidirectional links to your existing notes.
Morning reading, mid-day capture, evening review. Karpathy LLM Wiki fits into an Obsidian routine so your knowledge base grows on its own, without dedicated "wiki time".
Conversational queries create new knowledge. Karpathy LLM Wiki lets you save query results as wiki pages — the second path for your knowledge base to grow beyond source ingestion.
The schema/ folder defines how the plugin extracts knowledge — entity types, properties, alias rules, extraction templates. A deep dive into customizing your Obsidian knowledge base.
Knowledge from multiple sources sometimes conflicts. Karpathy LLM Wiki flags the contradiction, tracks both sides, and surfaces the conflict on the affected wiki page — strategies for resolving them in Obsidian.
v1.12.0 rearchitected the extraction prompt so Obsidian ingestion speed no longer scales with wiki size. The reasoning behind the change and how the speedup was measured.
Long-context models handle wikis differently. Karpathy LLM Wiki runs against DeepSeek, Gemini, Claude, GPT, Ollama, LM Studio — here is when each makes sense and what you'll pay.
How the codebase split a monolithic main.ts into a modular architecture — and added a streaming query interface on top without bloating the runtime.
Aliases, duplicates, dead links, orphans, empty pages — Karpathy LLM Wiki's Smart Fix All runs them in causal order so each repair doesn't undo the previous one.
Three automation layers keep your wiki current without manual upkeep. Karpathy LLM Wiki watches your Obsidian vault, runs lint on a schedule, and checks wiki health every time Obsidian starts.
v1.8.0 extended the UI from 2 languages to 8, with 269+ fields translated per locale. Inside the i18n process that keeps every release consistent — and what shipping to non-English users taught the team.
Wiki Output Language decouples your writing language from your reading language. Karpathy LLM Wiki generates the same content in 8 languages, independent of your Obsidian UI.
From an empty Obsidian vault to a living 100-page knowledge graph. Setup, your first sources, your first ingest — a beginner's walkthrough of Karpathy LLM Wiki.
The open-source Obsidian plugin that turns your notes into a self-organizing wiki. Karpathy LLM Wiki extracts entities, builds bidirectional links, and answers natural-language questions about your own knowledge.
The accumulation paradox: more notes mean harder access. Manual wiki maintenance scales worse than the value it adds — Karpathy LLM Wiki automates page creation, link writing, and updates in your Obsidian vault.
What a vault is, how Markdown works, why bidirectional links change how you take notes. The minimum knowledge you need before installing Karpathy LLM Wiki.
