About PageIndex
PageIndex is a document question answering engine that retrieves information by reasoning over a document's structure rather than by vector similarity search. The vendor describes it as a vectorless, reasoning based retrieval system that mirrors how a person reads a long document, and it cites 98.7 percent accuracy on the FinanceBench benchmark along with traceable, explainable retrieval.
The product is aimed at teams that need verifiable answers from long, complex documents such as financial filings, where knowing exactly which passage supported an answer matters. It is offered in several forms: a chat application for asking questions of uploaded documents, a developer dashboard and API for building retrieval into other software, and a demo booking route for organizations that want a walkthrough. The site also announces PageIndex Flash, described as a fast, open source tree indexing release.
PageIndex comes from a team of AI researchers with academic and industry backgrounds, including co founders of Humanloop, which was acquired by Anthropic.
The vendor does not publish a pricing page. Structured data on the site lists free entry points at zero cost, but paid plan prices and usage limits are not shown, so prospective customers should expect to use the dashboard or book a demo to learn about commercial terms.