When copy and paste is enough
For a short text-layer PDF with one simple table, paste a few rows and inspect the columns. A converter becomes useful when layouts repeat across pages, cells wrap, headers recur, or the table is embedded in a scan.
How extraction works
The single extraction contract identifies tables, headers, rows, source pages, and conservative cell types. Adjacent compatible tables can merge only when both sides show deterministic continuation evidence.
Exact repeated headers may be removed at a continuation boundary. Unrelated sections stay separate.
What it deliberately does not change
A visually merged cell, footnote, repeated page label, or wrapped line can still be misread. Page provenance and preview are there so you can compare the result with the original.
Insertion choices
Select the tables you want. Separate new app-owned tabs are the default. One small table can go into the current selection only if the full target is empty, unmerged, editable, and free of notes and validation. Source strings are inserted as literal text.
Common general-table questions
Will it preserve the PDF layout?
No. The output is table data for review, not a visual replica of the page.
Does it merge every table across pages?
No. Merging is conservative and adjacent-only. When continuation evidence is not exact, tables stay separate.
Can I download CSV?
Not in v1. The supported destination is the current Google spreadsheet.