InfraNodus is a text network visualisation and analysis tool that represents text as a knowledge graph. Words become nodes and their co-occurrences become connections, and graph-theory algorithms then identify clusters of related ideas (topic modelling), the most influential concepts and hidden intermediaries. Inputs include typed or pasted text, PDFs, Markdown, CSV and spreadsheet data, Google search and Scholar results, YouTube transcripts and comments, website URLs, RSS feeds and social posts. It is developed by Nodus Labs, which says the product has run since 2018 and that the research behind it goes back to 2011.
The AI layer sits on top of the graph. Gap detection finds structurally disconnected topic clusters, and LLM-generated insights and questions are aimed at brainstorming and research; the pricing page meters these as GPT-5 credits per hour, and the site also describes GraphRAG, which uses the knowledge graph to steer an LLM’s reasoning. Integrations include an MCP server for AI clients, an Obsidian graph-view plugin, Chrome and Firefox extensions and an API, and some of the tools underneath it are published as open source under the AGPL licence. The site describes user data as private by default.
Every plan starts with a free 14-day trial with no payment needed. Basic is EUR 12 a month for personal and academic use, with full graph analytics, the MCP server, the browser extensions and Obsidian plugin, 40 AI credits an hour and 1 MB uploads. Advanced is EUR 32 a month with API access, commercial use, dedicated support, 5 MB uploads and 80 GPT-5 credits an hour. Premium is EUR 66 a month with API integration support, an advanced MCP server, monthly training, 10 MB uploads (50 MB for PDFs) and 200 GPT-5 credits an hour. Annual billing saves up to 35%, prices exclude VAT for EU private customers, and student discounts and educational licences are available on request. The site names Greenpeace and Procter and Gamble among its users, which is vendor-cited.
