Upload a PDF
Extracted text is chunked, embedded, and indexed immediately.
API Key
PDF file
Source ID
(optional — defaults to the filename)
Domain
(optional)
Language
(optional)
Metadata
(optional JSON, merged into every chunk)
Must be a flat JSON object, e.g. {"unitId": "abc123"}. Leave blank for none.
Page range
(optional — leave both blank to index every page)
to
Use this to skip front matter (title/copyright/table of contents) and back matter (index) — the usual source of unhelpful chunks.
Upload & Index