PageIndexFrom the builder
Introducing PageIndex Flash
PageIndex introduced PageIndex Flash, a fast tree-indexing engine for long text-based PDFs that runs entirely on the user's machine, is open source and is now the default indexer in the SDK's local mode. PageIndex says Flash builds its hierarchical tree from the PDF's own layout instead of having a vision model infer the outline, which makes indexing faster, cheaper and more predictable. It runs no OCR, so it only suits text-based PDFs, and the company recommends PageIndex Cloud for scans and image-heavy files.
PageIndex Flash builds a hierarchical tree index by reading a PDF's own layout, rather than asking a vision model to infer the whole outline from scratch.
Saved . A shortened excerpt from the official blog; the original publisher retains authorship.