Choose a PDF
The file is read only in this browser. The owner must unlock encrypted or permission-protected PDFs first.
ImgIng / PDF to HTML
CLEAN-PAGE HTMLConvert PDF pages to one clean fixed-layout HTML
PDF is a page-description format, not ordinary reflowing web content. ImgIng reconstructs pages in PDF coordinates, supports Type 3 glyphs, inline and masked images, ICC and indexed colour, tiling patterns, axial and radial gradients and common soft masks, then identifies remaining review cases by page.
No PDF uploadType 3 vector glyphsMasks and axial / radial gradientsOne clean HTML
Capability quick facts
These facts describe the current product, not unshipped roadmap work.
- Where it runs
- In your browser; the PDF is not uploaded
- Default output
- Fixed-layout SVG, selectable text and one clean HTML
- Complex pages
- Type 3 glyphs, masked images, ICC / indexed colour, tiling patterns and axial / radial gradients
- Result UI
- Opens straight to pages with no zoom, report or original-PDF menu
Reviewed by the DataDance product and engineering team · Published · Updated
Finish in three steps
Confirm important settings before export; the processing location is always disclosed.
Rebuild pages locally
Recreate text, paths and images in PDF coordinates while reading ToUnicode, page rotation, crop boxes, links and usable fonts.
Preview, inspect and save
Review the fixed-layout result and complex-feature report, then use the system save dialog to export one HTML file.
Why not promise that every PDF becomes ordinary HTML with zero pixel differences?
PDF stores final drawing instructions. Ordinary HTML reflows with browser, font and viewport differences. Mesh shading, unusual blending, complex clipping, forms, 3D and multimedia also have no complete one-to-one HTML equivalent. ImgIng therefore uses fixed-layout SVG: page dimensions, text matrices, vector paths and image positions follow PDF coordinates while text remains selectable. Any non-equivalent feature is recorded in the workspace per-page report instead of being silently presented as a perfect conversion.
How are complex glyphs, colours, images and gradients reproduced?
ImgIng reads Type 3 CharProcs, caches each glyph as a reusable SVG vector symbol and adds an invisible selection layer. It handles ICCBased fallbacks, indexed palettes, inline images, one-bit ImageMask stencils, JPEG soft masks and common graphics-state soft masks. PatternType 1 tiling patterns, PatternType 2 axial shadings and ShadingType 3 radial shadings become inline SVG resources without multiplying the active graphics matrix twice.
How can everything fit in one clean HTML file?
Each page’s fixed-layout SVG, images, available fonts and necessary CSS are inlined. The export has no executable viewer runtime, carries no copy of the original PDF and adds no zoom, report, print or original-file toolbar. Open the file and the pages appear immediately. The quality report remains in the conversion workspace for source comparison before saving.
How are fonts kept faithful?
Every TrueType or OpenType font carried by the PDF and readable by the browser is inlined directly; font permission flags do not act as a visual-fidelity gate. When the PDF carries no font bytes, ImgIng first asks the device for the original family name. A system substitute is used only when the font is absent or its stream is technically unreadable by the browser, and the affected pages are reported. Type 3 fonts are rebuilt from their PDF vector glyph programs.
What is the security boundary?
The converter never executes PDF JavaScript, never turns PDF actions into web scripts and retains only http, https or mailto annotations after protocol checks. Output validation rejects external script, stylesheet, image or font loads, and the generated document includes a Content Security Policy.
Beta limits: this release focuses on fixed-layout reading, archiving, sharing and visual reproduction—not automatic responsive-article conversion or a drag-and-drop web editor. Isolated or knockout groups, masks with special transfer functions or matte premultiplication, CMYK screen colour, cross-browser JPEG 2000, mesh shading, text clipping, complex forms and multimedia can still require review; the workspace identifies the affected pages.
Updated 2026-10-10 · Live capability detection inside the tool is authoritative
Frequently asked questions
These visible answers match the current product behaviour and structured data.
Are the PDF and result uploaded?
No. Parsing, page reconstruction, font handling, preview and HTML packaging all happen locally.
Will the exported HTML open offline?
Yes. CSS, page SVG, images and browser-readable fonts are inlined; the result needs no ImgIng server and contains no viewer toolbar.
Why might a font not be embedded exactly?
If the PDF carries browser-readable font bytes, ImgIng always inlines them. Only an absent font or a technically unreadable font stream falls back to the original system family name or a compatible substitute, and the report identifies affected pages.
Explore ImgIng
Each task page documents real settings, limits and format advice—not keyword-swapped duplicates.