How Plot2Data Works

Start by uploading a graph image to Plot2Data. We support multiple image formats including PNG, JPG, SVG, and WebP. You can upload in three convenient ways: click the upload area to browse your files, drag and drop an image directly onto the page, or paste a screenshot from your clipboard using Ctrl+V (Cmd+V on Mac).

Plot2Data works with screenshots from any application, scanned pages from books or papers, saved graph images from websites, and exported figures from presentation slides. There are no file size limitations for typical graph images.

After uploading, you can optionally fine-tune the extraction settings to improve accuracy. By default, the AI automatically detects the graph type, axes, and data series without any manual configuration.

• X-Axis / Y-Axis Data Count: Specify the exact number of data points to extract along each axis. Set to 0 (default) for AI auto-detection.
• Error Bars: Toggle this option to extract upper and lower uncertainty ranges alongside data values. Essential for scientific and experimental data.
• Structural Mechanics Mode: Enable this for stress-strain curves to automatically calculate and extract Young's Modulus and ultimate strength values.

When you click "Extract Data," Plot2Data sends your graph image to Google Gemini Vision AI for analysis. The AI examines the image to identify graph elements including axes, labels, gridlines, data points, legends, and series markers. It determines the graph type automatically and extracts numerical values by mapping visual positions to their corresponding data values.

The extraction process typically takes 5 to 15 seconds depending on graph complexity. Multi-series graphs and graphs with dense data points may take slightly longer. The AI handles a wide variety of graph styles and visual designs.

Extracted data is displayed as an interactive table organized by data series. Each series includes its name and all associated data points with their X and Y values. You can copy the data to your clipboard with one click, or export it as a CSV file for use in Excel, Google Sheets, Python, R, MATLAB, or any other data analysis tool.

Supported Graph Types

Plot2Data supports a comprehensive range of graph types commonly found in academic papers, reports, and presentations:

See detailed examples with sample extracted data on our use cases page.

Tips for Best Results

• Use clear, high-resolution images. Graphs from high-DPI screens or 300+ DPI scans produce the most accurate results.
• Ensure axis labels and numbers are legible. The AI relies on readable text to establish value scales and data point positions.
• Crop your image to show only the graph. Remove surrounding text, page margins, and other non-graph elements for cleaner extraction.
• Include the legend for multi-series graphs. Visible legends help the AI correctly identify and separate different data series.
• Stick to standard 2D graph formats. Avoid heavily stylized 3D perspective views or artistic infographic designs, which may reduce extraction accuracy.
• Specify data point counts when known. If you know the exact number of data points on each axis, enter them in the settings to guide the AI.