about plotion

Turn any chart image into a dataset, not a screenshot.

Plotion exists for the researcher who found the exact number they need buried in a PDF figure, with no underlying data available anywhere. Instead of eyeballing pixels in an image editor, you calibrate the axes once and click the points, and Plotion gives you real numbers, with a computed margin of error, ready to cite.

How it works

Upload a chart image, tell Plotion what type of chart it is and what the axes mean, mark the data points, and export. Five steps: Upload, Setup, Calibrate, Pick, Export. Nothing about the process requires a second tool.

Your chart image never leaves your browser

Digitizing happens entirely client-side. The image you upload, the calibration you set, and every point you place stay in your browser's local storage, not on our servers. Only the data you choose to export (CSV, JSON, or XLSX) ever leaves your machine, and that's your call.

Seven chart families, one workflow

Each chart type gets its own calibration model instead of a one-size-fits-all fit, so the math matches how the chart actually encodes its values.

  • XY (scatter + line)
  • Bar (grouped / stacked)
  • Pie / donut
  • Polar
  • Ternary
  • Geo
  • Scale-bar

Built for people who cite their sources

Every extracted dataset ships with a margin of error computed from the actual image and your specific axis calibration, not a static estimate. If you're going to put a number in a paper, you should know how much to trust it.