Snip Snipping Tool Chrome Extension OCR API Files API Secure Conversion Service
Make Documents Accessible Process Chemical Documents Collaborate on Documents Developer Solutions Train Language Models Support Academic Research Artificial Intelligence Fintech Edtech Pharma & Chemical Universities & Schools
Handwriting Recognition Digital Ink On-prem PDF Cloud Mathpix Markdown All Supported Languages Image Conversion PDF Conversion Markdown Conversion Table OCR Mathpix CLI PDF Search PDF Reader PDF Data Extraction Chrome Extension View Conversion Gallery
Snip APIs SCS
Mobile Desktop Web Chrome Extension
Mathpix Snip Apps Mathpix OCR API Mathpix Markdown Python SDK
Blog
About Careers Contact
Contact Get Started

Playground

The Playground tab on console.mathpix.com runs real API requests from your browser. It is the fastest way to see what Mathpix returns for a document of your own before you write any code, and the fastest way to check a suspect file when output does not look the way you expected.
Requests made in the Playground use a real API key and count towards that key’s usage and your bill, exactly like requests from your own code.

What you can run

The Playground covers the two endpoints most integrations start with:
  • v3/text — send an image and get back text, with math, tables, and diagrams recognized in place.
  • v3/pdf — send a document and get back both its text data and its segmentation coordinates, so you can see how Mathpix breaks the page into elements.
Accepted uploads are .pdf, .png, .jpg, and .jpeg. For the full range of input formats the API accepts, see Supported Formats.

Running a request

  1. Upload a file. The endpoint switches automatically — a PDF selects v3/pdf, an image selects v3/text — and you can override it from the Endpoint dropdown.
  2. Pick an API key. Only enabled keys appear in the list. Whichever key you choose is the one that gets billed and the one the request shows up under in Usage and results.
  3. Pick a region. You can send the request to the global endpoint or to the EU endpoint, which is a useful way to confirm EU data residency behaves the way you expect before you repoint production. See Server locations.
  4. Click Convert.
For documents, the Playground shows a live conversion status and percent complete while the job runs, and an Update conversion status button to re-check on demand. Large documents can take several minutes; the page tells you when a file is still converting rather than failing silently.

Reading the output

The response pane shows the raw JSON the API returned — the same payload your code would receive, so you can copy a field path straight out of it.
For documents, the Playground also overlays bounding boxes on the rendered pages, drawn from the line-level data at v3/pdf/{pdf_id}.lines.json. Each box is one element Mathpix detected. Use Hide bounding boxes to compare the clean render against the segmentation. This view answers a specific question quickly: when a heading, table cell, or equation comes out wrong, the boxes show whether the model mis-segmented the page or mis-read a correctly segmented element.
Document requests in the Playground are submitted with conversion_formats of docx and tex.zip, math_inline_delimiters of $…$, and rm_spaces enabled. Your own requests can set any of the request parameters — the Playground shows you the shape of the response, not the full parameter surface.

When to use something else

  • To explore request parameters, use the API reference or the Quickstart — the Playground uses a fixed set of options.
  • To look at requests you already sent from your own code, use the Results tab — see Usage and results.
  • To process many files, use files/v1. The Playground is one file at a time.