unplot

vector PDF

Read curves straight off a vector PDF — the actual numbers behind the lines, calibrated, with overlapping ones pulled apart. Runs entirely in your browser; nothing is uploaded.

Works on vector PDFs (real drawn paths) — this in-browser build doesn't read scans. The Python package is more capable and handles raster / scanned plots too.

Source PDF
reading the PDF…
Multi-plot page? Drag a box around one plot's axes (tick labels just outside) to extract only that plot.
Extracted data re-drawn from the numbers · hover a point
CurvePointsPeakConfidenceViolationMonotonicStatus