Built for the people who actually pull data from charts
Three audiences who otherwise spend hours redrawing or estimating chart values by eye. DataFromChart cuts that to under a minute.
<1 min
Average extract time
Pixel-precise
On linear, log, and categorical axes
Up to 10
Series in one pass
CSV / JSON
Or paste straight into Excel
Three teams, three deep-dive guides
Pull data out of any published figure
Extract numerical data from journal charts, technical reports, and conference papers — without redrawing or estimating by eye.
Read more → For consultantsPull the data out of any deck chart
Stop redrawing slides in PowerPoint to get at the underlying numbers. Drop the screenshot in, paste straight into Excel.
Read more → For financial analystsReconstruct any chart from a research report
Pull quarterly figures, earnings projections, and sector trends out of analyst PDFs. Screenshot to CSV in under a minute.
Read more →From "redraw it by eye" to "drop it in"
Eyeballed values off a published figure, recreated it in Excel, hoped peer review didn't catch the inaccuracy.
Open the screenshot, drop points on the data, calibrate axes from the tick labels. Export CSV with pixel-level precision.
Spent the morning redrawing the client's chart in PowerPoint because the underlying spreadsheet was lost three quarters ago.
Snap a screenshot of the deck, extract the series in under a minute, paste straight into the new model.
Took a ruler to a PDF chart of quarterly EPS, transcribed values by hand, fudged the years that were too close to read.
Calibrate axes by dragging reference lines onto the tick labels, then click each data point. Multi-series in one pass. AI auto-extraction is available now — 3 free runs per account, then $5 for 20.
Everything the source chart contains
DataFromChart handles the chart types and axis shapes that actually appear in research papers, decks, and analyst reports.
Scatter, line, bar, column, area
Five chart types covered out of the box. Combo charts work — extract each series separately and merge.
Linear, log, and categorical axes
Log-scale axes are calibrated by entering two reference values — DataFromChart handles the math.
AI auto-extract
Opt in on the upload step and one pass classifies the chart, calibrates the axes, and plots the points. Refine anything, then export.
Multiple series in one pass
Color-tag each series, label them, extract every line on the chart without starting over.
Zoom and pan for precision
Tightly-packed data points need a 4× zoom and a steady hand. The canvas supports both.
CSV, JSON, or clipboard
Export in the format your next tool expects. Or just copy to clipboard and paste into Excel.
Common questions
Do I need to install anything?
No. DataFromChart runs entirely in the browser. Open a chart, calibrate, export — no install, no plugin.
What chart types are supported?
Scatter, line, bar, column, area, and combinations of those. Linear, log, and categorical axes are all handled.
How accurate is the extraction?
Pixel-level. Accuracy is bounded by the resolution of the source chart and the calibration you provide for the axes.
Can I extract more than one data series at once?
Yes. Each series gets its own color and label, and DataFromChart keeps them separate in the export.