What happened

On 27 September 2026, the PDF Association announced version 1.2 of its Deriving PDF to HTML algorithm. The document is guidance for developers deriving structured HTML from well-tagged PDF, especially material aligned with Well-Tagged PDF (WTPDF). It is a technical method, not a promise that any arbitrary PDF can be turned into a good web page.

The association compares version 1.2 with the published 1.0 release. Version 1.1 was an internal intermediate and was never published. That history matters when evaluating claims about a supposed 1.1 release or treating the new number as evidence that a consumer converter has changed.

How the algorithm matters

A page screenshot reproduces appearance but does not tell a browser which text is a heading, list, link or form label. The revised guidance puts more emphasis on mapping PDF structure to meaningful HTML, using namespaces and role information appropriately, and handling page breaks and page labels. It also covers metadata such as language and title, which help readers navigate a document.

Consider a report with a table of contents and a form. A useful HTML version needs links that go to the right sections, fields with understandable labels, and an order that works on a narrow screen. It is not enough for the pixels to resemble the PDF. WTPDF gives implementers a more defined source structure on which to base those decisions.

What it means for users

If you publish documents online, ask for an accessible HTML source from the original authoring workflow when possible. When a vendor offers PDF-to-HTML conversion, examine headings, lists, links, form controls and reading order in the actual output. Check on mobile and with keyboard and assistive technology, especially for public forms or instructions.

Convertios does not offer PDF-to-HTML conversion. For a different task, you can convert a text-based PDF to Word to work on its wording, or export a PDF page as JPG for a visual reference. A JPG is a picture, not a structured HTML document. The related PDF/UA-1 notes show why structure also matters inside PDF.

Practical checklist

  • Keep the original editable source and use its web export when it preserves semantics.
  • If deriving HTML from PDF, inspect heading levels, lists, links, form labels and page navigation.
  • Test the output on narrow screens and with keyboard and assistive technology.
  • Compare the HTML with the PDF for omitted or reordered content before publication.

Limitations and what has not changed

The PDF Association explicitly cautions that derived HTML does not automatically inherit the accessibility characteristics of the source PDF. Poorly tagged or image-only PDFs offer less usable structure to derive. The algorithm does not repair every authoring mistake, perform OCR, or add a PDF-to-HTML tool to Convertios. Treat any output as a document to review, rather than as an accessibility certificate.

Primary sources