AI Knowledge Graph turns unstructured text into interactive knowledge graphs. Feed it a document, and an LLM of your choice extracts the people, organizations, places, events, and concepts inside it, along with the relationships between them, rendered as an explorable graph you can click through, question, and serve on the web.
This open-source project lives on GitHub, where it has earned more than 3,100 stars under the Apache-2.0 license, and it remains actively developed, with updates in September 2026. The system reads plain text, Markdown, reStructuredText, PDF, and Word documents, in any language, splitting large files on sentence boundaries and extracting in parallel. The LLM returns typed Subject-Predicate-Object triples: every entity is classified (person, organization, place, event, technology, product, work, date, or concept), and every relationship keeps the exact sentence it came from, so the graph stays traceable to its sources instead of becoming an unverifiable blob. Entity standardization merges case, stop-word, and plural variants, with an optional LLM pass for the rest, and conservative inference passes connect isolated clusters with well-known relationships. Companion commands let you chat with the finished graph and ask it questions, or serve everything through a local web interface. Live examples on the project site, from the Industrial Revolution to the Apollo program, show the output without installing anything.
AI Knowledge Graph is free and open source under the Apache-2.0 license. You bring your own LLM API key (or local model), so your only cost is the model usage itself.
It suits researchers, analysts, journalists, and knowledge engineers who need to see the structure hidden inside long documents, as well as RAG builders who want graph-shaped context extracted from raw text with sources attached.