Free chart maker

How to use it

1. Paste your table, open a CSV or JSON file, or load a sample. The first row should hold column names.
2. Check the columns: pick the label column, the number columns to plot, and optionally a column to split into series.
3. Take the suggested chart type or choose another, then add a title, axis titles and colors.
4. Download PNG, SVG or PDF, or copy the image URL or Markdown to embed it.

What you can make

• Grouped or stacked bar charts, vertical or horizontal
• Line and area charts with up to 12 series
• Pie and doughnut charts with automatic percentages
• Value labels, $ or % prefixes and suffixes, axis titles, a subtitle for the source

Worked example: long-format data

Exports from analytics tools often have one row per date and category, like Quarter, Region, Sales. Load the "Sales by region" sample: the maker spots that Region repeats, splits it into one series per region, sums repeated rows, and keeps quarters in time order. You get a grouped bar chart with four regions without reshaping anything in a spreadsheet.

How the maker reads your data

• Numbers like $1,200, 12%, 1e3 and (300) for negatives are understood.
• Dates such as 2026-01-15, 1/15/2026, Jan 2026 and 2026 Q1 are recognized and sorted in time order.
• Empty cells and values like N/A are plotted as 0, and the count is shown under the preview.
• Errors point at the line: an unclosed quote or a JSON typo says which line to fix.

Your chart as a link

Every chart you make here has a permanent image URL on render.cuelark.com. Paste it into a README, a Notion page, an email or a Markdown doc and the image renders there, so you don't upload anything. The URL holds the chart data itself, so anyone with the link can see the numbers. Don't put private data in a shared chart link.

Charts from code or an AI assistant

The same renderer runs behind the Chart image API and the MCP server. Your scripts, spreadsheets, n8n flows or Claude and ChatGPT can make the identical chart from JSON.

Do it from code or an AI assistant

The same renderer is an HTTP API and an MCP tool, so this exact result can come from a script or from Claude, ChatGPT or Cursor. See the Render API reference and the MCP setup guide.

HTTP API

# GET: the URL is the image (embed it in Markdown, email or Slack)
curl -o chart.png "https://render.cuelark.com/v1/chart?c=%7B%22type%22%3A%22bar%22%2C%22title%22%3A%22Sales%20by%20month%22%2C%22labels%22%3A%5B%22Jan%22%2C%22Feb%22%2C%22Mar%22%2C%22Apr%22%2C%22May%22%2C%22Jun%22%5D%2C%22datasets%22%3A%5B%7B%22label%22%3A%222025%22%2C%22data%22%3A%5B42%2C48%2C51%2C60%2C58%2C71%5D%7D%2C%7B%22label%22%3A%222026%22%2C%22data%22%3A%5B50%2C57%2C66%2C74%2C80%2C92%5D%7D%5D%7D"

# POST: same chart at 2x for slides
curl -X POST https://render.cuelark.com/v1/chart -H "content-type: application/json" \
  -d '{"chart":{"type":"bar","title":"Sales by month","labels":["Jan","Feb","Mar","Apr","May","Jun"],"datasets":[{"label":"2025","data":[42,48,51,60,58,71]},{"label":"2026","data":[50,57,66,74,80,92]}]},"format":"png","scale":2}' -o chart@2x.png

MCP tool call: render_chart

{
  "method": "tools/call",
  "params": {
    "name": "render_chart",
    "arguments": {
      "type": "bar",
      "title": "Sales by month",
      "labels": [
        "Jan",
        "Feb",
        "Mar",
        "Apr",
        "May",
        "Jun"
      ],
      "datasets": [
        {
          "label": "2025",
          "data": [
            42,
            48,
            51,
            60,
            58,
            71
          ]
        },
        {
          "label": "2026",
          "data": [
            50,
            57,
            66,
            74,
            80,
            92
          ]
        }
      ]
    }
  }
}

Frequently asked questions

Related free tools from Cuelark

• Percentage calculator