Loading Techie Utils…

PDF to JSON Converter

Export a PDF's text as structured JSON. Free, and runs entirely in your browser — nothing you enter is uploaded, stored or sent to a server.

About PDF to JSON

A PDF's text as JSON, in one of two shapes. Blocks gives headings and paragraphs — readable structure, inferred the same way the Markdown conversion infers it. Text runs gives every run of text with the coordinates it was drawn at, which is the raw material.

The second shape is the interesting one for anything doing its own analysis. A PDF's text is stored as runs positioned on the page, and every layout inference any tool makes — paragraphs, columns, tables, reading order — is computed from exactly those coordinates. Getting them directly means you can apply your own rules instead of accepting somebody else's.

Both shapes are page-oriented, because the page is the only division a PDF genuinely has.

Where it helps

  • Two shapes: inferred structure, or the raw positioned runs behind it.
  • Page dimensions are included, so coordinates can be interpreted without opening the file again.
  • Straightforward JSON — no schema to learn, nothing nested deeper than it needs to be.
  • The document stays in your browser, which matters for anything being prepared for a model or a pipeline.
  • Offered as a download as well as copyable text.

Limits

  • The runs shape is verbose: a hundred pages can be several megabytes of JSON.
  • Coordinates are in PDF points with the origin at the bottom-left, which is upside down relative to most graphics code.
  • No font details beyond the name, and no colour information.
  • Images, annotations and form values are not included — only text.
  • A scan contains no text, and produces an empty result rather than an error.

Frequently asked questions

Which shape should I use?

Blocks if you want the text and its structure — an index, a summary, a model's context. Runs if you are doing your own layout work, because it gives you what every inference is computed from rather than the conclusion.

Why is y measured from the bottom?

Because that is PDF's coordinate system: the origin is the bottom-left of the page and y increases upwards. Subtract from the page height to get a top-down coordinate.

Can I get the images or the form values too?

Not here — Extract Images and the form tools cover those. This is text.

Is this suitable for RAG?

The blocks shape is a reasonable starting point: each page's headings and paragraphs, ready to chunk. Whether it is good enough depends entirely on your documents' layout, which is worth checking on a handful before running a library through it.

Related tools

More Images & Files tools