Paste rows of data as CSV, JSON, or one "Label, Value" per line, pick a chart type, and draw it right in your browser - no upload and no account.
On-device AI insight (optional)
Recent charts
Data Visualizer
Turn rows of data into a chart you can read at a glance - paste your numbers, choose a chart type, and the tool plots them for you on the page.
What you can do
- Switch between bar, line, pie, doughnut, polar area, radar, scatter, bubble, and stacked-bar chart types to find the clearest view of your numbers.
- Load a built-in sample dataset - monthly sales, product comparison, or expense categories - to start with one click.
- Give the chart a title, and it redraws as you change the type, title, or data.
- Reopen an earlier chart with a single click - your recent charts are remembered in this browser.
Three ways to enter your data
Paste CSV, JSON, or one "Label, Value" pair per line - Data Visualizer detects the format for you and draws the chart in this browser.
| Input format | What it looks like |
|---|---|
| CSV | A header row, then comma-separated rows - one column becomes the labels and the rest become the values. |
| JSON | An array of objects (a label key plus numeric values), or a single key and value object. |
| Label, Value | One "Label, Value" pair per line - the simplest way to chart a short list. |
Optional on-device AI insight
On a WebGPU browser you can run an optional private AI insight that reads your data on your device and rebuilds the chart - the model runs locally and nothing is uploaded. Without WebGPU, the chart is still built for you by the deterministic engine.
Clicking "Get AI insight" shows a loading percentage while the model prepares, since nothing runs until it is ready. If the model's answer cannot be turned into a valid chart, the tool does not fail or show an error - it keeps the deterministic chart already on the page, so you are never left without a result. On a browser without WebGPU the step says WebGPU is missing and points out that the chart above already works without it.
Your rows become a chart without leaving the tab
The whole pipeline runs in this browser: the parser reads your pasted CSV, JSON, or label-value rows in the page, and the chart engine draws every bar, line, and slice onto an HTML5 canvas on your device. There is no upload and no account, so the data you paste never touches a server. The recent charts you reopen are remembered in this browser's local storage - kept on this device, never synced anywhere else.
The chart lives on the page rather than downloading as a file, so to keep or share one, capture it with your device's screenshot tool. If you would rather pass on the underlying numbers, copy them straight from the box you pasted them into.
This is one of the developer tools. If your numbers arrive as a JSON payload, validate and inspect them first with the JSON Parser, or browse the full set of formatters, minifiers, and converters on the hub.
Read the guide: How to Visualize Data Online - Make a Chart From Your Numbers
Frequently Asked Questions
What does Data Visualizer do?
Turn rows of data into a chart you can read at a glance.
When should I reach for data visualizer?
Switch between chart types to find the clearest view of your numbers.
What complementary tools work well alongside data visualizer?
For exporting or converting the underlying data, see the related developer tools.
Can I upload a CSV file instead of pasting it?
No - there is no file-upload button. Copy your CSV, JSON, or Label,Value rows and paste them into the data box, then pick a chart type.
Can Data Visualizer pull data from Google Sheets or a database?
No - it only works with data you paste in or one of the built-in sample datasets. It does not connect to Google Sheets, a database, or any outside API.
Does the on-device AI insight need a download?
Yes, the first time you use it in a given browser - the AI model downloads to that browser before it can read your data. On a browser without WebGPU support, the AI insight option is not available and the deterministic chart still builds normally.