Question 1

Will the HTML look exactly like my PDF?

Accepted Answer

Pick your trade-off, we don't pretend there is only one answer. Flow mode produces a clean, responsive, readable web page — text, paragraphs, two heading tiers, bold/italic and lists survive, but positions, columns, images and tables do not. Layout mode rebuilds every page at its exact pixel size with lines placed at their real PDF coordinates (pt × 96/72), so it looks like the pages — but the text never rewraps on small screens, and the baseline placement uses a standard 0.8 em ascent approximation. Neither mode draws images or vector art.

Question 2

How do headings, bold, italic and lists work?

Accepted Answer

Straight from the text layer: the document's body size is the most-used font size; the two clearly-larger size clusters become

and

(a third "tier" stays body text — we don't invent deeper headings). Bold and italic runs are recovered by matching pdf.js's internal font ids with the PDF's own /Font table, so "Helvetica-Bold" text becomes and "Helvetica-Oblique" becomes . Lines starting with •, ·, - or "1." become real
/
list items; heading-sized lines are never mistaken for list items.

Question 3

What about Hindi and other scripts?

Accepted Answer

This is where HTML beats PDF outright. A PDF must embed font bytes to render Devanagari; an HTML file just carries plain UTF-8 and every browser already ships system fonts for it. So हिंदी text converts with no font file, stays selectable, searchable and copyable, and the output stays tiny. The file's lang attribute is even set from the content (mostly-Devanagari documents get lang="hi").

Question 4

Where is my file processed, and what are the limits?

Accepted Answer

Entirely in this browser — a Web Worker reads the PDF with the self-hosted pdf.js, extracts and rebuilds the text, and writes the single .html locally. Nothing is uploaded, no sign-up, no watermark. One PDF at a time; page ranges like 1-3, 5, 8- convert only the pages you pick. Very large documents are capped honestly: 2,000 blocks in flow mode and 6,000 positioned lines in layout mode. A PDF with no selectable text anywhere (a pure scan) stops with a clear message instead of an empty file — run the OCR tool on it first, then convert the result.

PDF to HTML

How it works

Frequently asked

PDF to HTML converter