About Venz

Venz makes it easy to create a chart from numbers.

Venz visualizes numeric data using SVG charts. You feed it numbers and it tries to detect the data structure automatically. The chart can then be downloaded in SVG, PNG, webP or AVIF format. Or create a link that contains all details to share.

There are multiple ways to create a chart:

  • By pasting text
  • By dropping a file onto the chart
  • By creating a URL

The numbers can be in plain text or JSON format.

Contents

Unlabeled numbers

Numbers separated by space, comma, semicolon, or newline:

84, 4, 80, 60, 17, 44, 64, 10, 86, 4
77
98
59
90
73
55
77
47
53
12

Copy-paste the numbers onto the chart, or try this example link:

https://try.venz.dev?type=line&data=84*4*80*60*17*44*64*10*86*4

Paste numbers repeatedly to add series, or try this example link:

https://try.venz.dev?type=line&data=84*4*80*60*17*44*64*10*86*4&data=77*98*59*90*73*55*77*47*53*12

Labeled numbers in rows

2025-01 150
2025-02 120
2025-03 180
2025-04 170
2025-01 150 150 160
2025-02 120 110 120
2025-03 180 180 210
2025-04 170 190 180
1.0.0 3
1.0.1 3.05
1.1.0 2.88
1.2.0 2.9
1.2.1 2.8

If the labels are recognized as date or semver, they're sorted accordingly.

Copy-paste the labeled numbers onto the chart, or try this example link:

https://try.venz.dev?type=line&legend=br&zero=0&row=2025-01&row=2025-02&row=2025-03&row=2025-04&data=150*150*160&data=120*110*120&data=180*180*210&data=170*190*180

Pasting additional data results onto the chart results in merged data with additional series.

Labeled numbers in columns

col-A col-B col-C
----- ----- -----
 1.2    2    2.5
 1.1    1.9  2.4
1.15    1.8  2.4
0.98    2.2  2.6
col-A,col-B,col-C
1.2,2,2.5
1.1,1.9,2.4
1.15,1.8,2.4
0.98,2.2,2.6

Copy-paste the labeled numbers onto the chart, or try this example link:

https://try.venz.dev?type=line&legend=br&row=Col-A&row=Col-B&row=Col-C&data=1.2*1.1*1.15*0.98&data=2*1.9*1.8*2.2&data=2.5*2.4*2.4*2.6

Instead of separate parameters, the data can be the full text:

https://try.venz.dev?type=line&legend=br&data=col-A+col-B+col-C +1.2++++2++++2.5 +1.1++++1.9++2.4 1.15++++1.8++2.4 0.98++++2.2++2.6

JSON with numbers

[5, 4, 5]
[
  [5, 4, 6],
  [4, 5, 3],
  [3, 6, 4]
]

Copy-paste the labeled numbers onto the chart, or try an example link:

https://try.venz.dev?type=line&data=[5,3,4]

https://try.venz.dev?type=line&data=[[5,4,6],[4,5,3],[3,6,4]]

JSON with labels and numbers

[
  ["Bananas", 5],
  ["Apples", 4],
  ["Oranges", 3]
]

Copy-paste the labeled numbers onto the chart, or try this example link:

https://try.venz.dev?type=bar&labelY=Units&data=[["Bananas",5],["Apples",4],["Oranges",3]]

JSON with labels and numbers series

[
  ["Bananas", [5, 4, 5]],
  ["Apples", [4, 5, 4]],
  ["Oranges", [3, 6, 3]]
]

Copy-paste the labeled numbers onto the chart, or try this example link:

https://try.venz.dev?type=line&data=[["Bananas",[5,4,6]],["Apples",[4,5,3]],["Oranges",[3,6,4]]]

Hyperfine

Venz was originally created to visualize hyperfine output. The name Venz is derived from bENchmark VisualiZer.

Drop the file or paste its contents generated by hyperfine --export-json. See the examples and use configurations to create and manage hyperfine commands.

Example command:

hyperfine --warmup 3
  'sleep 0.35'
  'sleep 0.31'
  'sleep 0.33'
  --export-json results.json

Example output:

{
  "results": [
    {
      "command": "sleep 0.35",
      "mean": 0.35,
      "stddev": 0,
      "median": 0.35,
      "user": 0.0001,
      "system": 0.0001,
      "min": 0.35,
      "max": 0.35,
      "times": [0.35, 0.35, 0.35],
      "exit_codes": [0, 0, 0]
    },
    {
      "command": "sleep 0.31",
      "mean": 0.31,
      "stddev": 0,
      "median": 0.31,
      "user": 0.0001,
      "system": 0.0001,
      "min": 0.31,
      "max": 0.31,
      "times": [0.31, 0.31, 0.31],
      "exit_codes": [0, 0, 0]
    }
  ]
}

Drop the file results.json or copy-paste its contents onto the chart, or try this example link:

https://try.venz.dev?type=median&zero=0&data={"results":[{"command":"sleep+0.35","mean":0.36833403254,"stddev":0.0022131087112131635,"median":0.36842083684000004,"user":0.004151359999999998,"system":0.0044277399999999995,"min":0.36323077384,"max":0.37231769084000005,"times":[0.36811944084000003,0.36773835784000003,0.37231769084000005,0.36323077384,0.36910831584000003,0.36815719084000004,0.36806119084000005,0.36868448284000005,0.36923681584,0.36868606584],"exit_codes":[0,0,0,0,0,0,0,0,0,0]},{"command":"sleep+0.31","mean":0.32774909924,"stddev":0.0016290075685424868,"median":0.32849373234,"user":0.00461136,"system":0.004440939999999999,"min":0.32424739884000003,"max":0.32908635784,"times":[0.32766448284000005,0.32852669084,0.32894244084,0.32859089984,0.32424739884000003,0.32908635784,0.32780769084000005,0.32539560684,0.32846077384000005,0.32876864984000004],"exit_codes":[0,0,0,0,0,0,0,0,0,0]},{"command":"sleep+0.33","mean":0.34931401164000003,"stddev":0.0038889942457800007,"median":0.34879402384,"user":0.0042200599999999994,"system":0.004495439999999999,"min":0.34237498284,"max":0.35686323184,"times":[0.34762089884,0.34802077484000005,0.34890223184,0.34985627484000004,0.34892852384000006,0.34237498284,0.34777194084,0.34868581584,0.35686323184,0.35411544084],"exit_codes":[0,0,0,0,0,0,0,0,0,0]}]}

Mitata

Drop the file or paste its contents generated by run({ format: 'json' }). Series data is truncated to 100k samples each (a notification will show if there is any data loss).

Example script:

import { run, bench } from 'mitata';

bench('sleep 2.1s', () => new Promise(resolve => setTimeout(resolve, 2100)));
bench('sleep 2.2s', () => new Promise(resolve => setTimeout(resolve, 2200)));
bench('sleep 2.3s', () => new Promise(resolve => setTimeout(resolve, 2300)));

await run({ format: 'json' });
node sleep.ts > results.json

Example output:

{
  "layout": [{ "name": null, "types": [] }],
  "benchmarks": [
    {
      "alias": "sleep 2.1s",
      "runs": [
        {
          "stats": {
            "samples": [
              2100617875, 2101011375, 2101379667, 2101386334, 2101409917,
              2101517584, 2101943916, 2102111417, 2102819000, 2102983959,
              2103255958, 2107508833
            ],
            "min": 2100617875,
            "max": 2107508833,
            "p25": 2101379667,
            "p50": 2101517584,
            "p75": 2102819000,
            "p99": 2103255958,
            "p999": 2103255958,
            "avg": 2102328819.5833333,
            "ticks": 12
          },
          "args": {},
          "name": "sleep 2.1s"
        }
      ],
    }
  ]
}

Drop the file results.json or copy-paste its contents onto the chart, or try this example link:

https://try.venz.dev?type=median&zero=0&data={"layout":[{"name":null,"types":[]},{"name":null,"types":["s"]},{"name":null,"types":[]}],"benchmarks":[{"alias":"sleep+2.1s","runs":[{"stats":{"samples":[2100617875,2101011375,2101379667,2101386334,2101409917,2101517584,2101943916,2102111417,2102819000,2102983959,2103255958,2107508833]},"name":"sleep+2.1s"}]},{"alias":"sleep+2.2s","runs":[{"stats":{"kind":"fn","samples":[2200299791,2200441500.000004,2200454208,2201334625,2201374125,2201430041,2201702959,2201707625,2201790375,2202938000,2203260584,2206625291]},"name":"sleep+2.2s"}]},{"alias":"sleep+2.3s","runs":[{"stats":{"kind":"fn","samples":[2299657708,2300495375,2300578541,2300587458,2300834125,2300846292,2301183417,2301186208,2301333500,2301397208.9999924,2301521375,2301543250],"min":2299657708,"max":2301543250,"p25":2300578541,"p50":2300846292,"p75":2301333500,"p99":2301521375,"p999":2301521375,"avg":2300930371.4999995,"ticks":12,"heap":{"_":11,"total":22904,"min":1928,"max":2240,"avg":2082.181818181818}},"name":"sleep+2.3s"}]}]}

Create Chart URL

Create links to share visualizations of your data.

Example: https://try.venz.dev?type=line&data=3*2*3*4*2*5&data=4*3*2*5*6*4

This describes version 2 of the URL format. URLs with a version 1 parameter redirect to their version 2 equivalent.

URL parameters

  • data
    • numbers of one row
    • separated by *, comma, semicolon or whitespace (* needs no escaping in URLs)
    • can be repeated
    • a single data starting with { or [ is read as JSON (e.g. Hyperfine output)
  • row
    • name of a row
    • one name per parameter
    • can be repeated
    • first row belongs to first data, etc.
    • only used when there are as many row as data parameters
  • col
    • names of the columns: the positions within each row (e.g. versions, dates or sizes)
    • can be repeated
    • default: Series 1, Series 2, etc.
  • type
    • chart type
    • default: median
    • values: median, box, scatter, line, bar
  • by
    • one line, bar or box per row or per column, each with a legend entry and color (see series and samples)
    • default: row
    • values: row, col
  • samples
    • whether the values of each line, bar or box are samples (see series and samples)
    • default: 1 when the other dimension has no names (no col with by=row, no row with by=col), otherwise 0
    • values: 0, 1
  • labelY
    • label for y-axis
    • default: median, or median (unit) with unit
  • labelX
    • label for x-axis
    • default: sample # when the x-axis shows sample numbers, version or date when it shows row names that look like versions or dates, otherwise empty
  • color
    • CSS color
    • hex colors work without # (color=ff8800), or encode # as %23
    • can be repeated
    • first color belongs to first row and to first column, etc. (whichever is in the legend)
  • legend
    • legend position (inside the plot area)
    • default: tr
    • values: tr (top right), tl (top left), br (bottom right), bl (bottom left), none
  • zero
    • 0 starts the y-axis near the lowest value instead of at zero
    • default: 1
  • unit
    • unit of the numbers, the y-axis then shows the best fitting unit (ns, µs, ms or s)
    • values: s, ns
  • command
    • command line of a row (e.g. from Hyperfine), shown in the series list
    • can be repeated
  • embed
    • 1 shows only the chart, without header and controls (e.g. for iframes)
  • theme
    • color theme, applies together with embed=1
    • values: dark, light, high-contrast

Invalid values fall back to the default. Links created with the share button are capped at 15,000 characters, longer series get truncated.

Series and samples

by decides what gets a line, bar or box, a legend entry and a color: each row, or each column. The other dimension is either shown by name on the x-axis, or holds samples (such as repeated measurements):

bysamplesline, scatterbox, median, bar
row0a line per row, across the col namesa mark per row at each col name
row1a line per row, across sample numbersa mark per row, summarizing its values
col0a line per column, across the row namesa mark per column at each row name
col1a line per column, across sample numbersa mark per column, summarizing its values

So without by or samples, a line chart of rows with col names draws a line per row across those names: https://try.venz.dev?type=line&labelY=duration+(s)&row=astro&row=query&col=6.0.0&col=6.5.0&col=6.44.2&data=2.10*1.57*1.66&data=1.91*1.99*2.09

And a box plot of rows without col names draws a box per row: https://try.venz.dev?type=box&labelY=duration+(s)&row=astro&row=query&data=2.1*2.2*2*2.3&data=1.9*2*1.8*2.1

Grouped bars, with the rows on the x-axis and a bar per column: https://try.venz.dev?type=bar&by=col&labelY=duration+(s)&row=astro&row=query&col=v5.88.0&col=v6.0.0&data=4*2&data=3.8*1.7

Example code snippets to create URL

const url = new URL('/', 'https://try.venz.dev');
url.searchParams.set('type', 'line');
url.searchParams.append('data', '1*2*3');
url.searchParams.append('data', '4*5*6');

console.log(url.toString());
const url = new URL('/', 'https://try.venz.dev');
url.searchParams.set('type', 'line');

const series = { a: [1, 2, 3], b: [4, 5, 6] };
for (const [name, values] of Object.entries(series)) {
  url.searchParams.append('row', name);
  url.searchParams.append('data', values.join('*'));
}

console.log(url.toString());

Version 1

URLs with any of the v1 parameters below, or with type=pivot, use version 1 of the format. Chart page and image URLs redirect them permanently (301) to their version 2 equivalent, which renders the same chart. Version 2 does not use these parameter names.

v1v2
labelrow
lcol
lx, lylabelX, labelY
pivot, transposeby, samples (see below)
lp, lp=nlegend, legend=none
br=1zero=0
chrome=0embed=1
ctnone (it had no effect)
type=pivottype=line&by=col&samples=0
pivottransposev2
00by=row&samples=1
10by=col&samples=0
01by=col&samples=1
11by=row&samples=0

In v1, pivot defaults to 0 for line and scatter charts and to 1 for the other chart types. v1 also ignores labelX, labelY, unit, color, command and l unless the rows are labeled.

Create Image URL

Generate chart images directly via URL. Append chart parameters as query strings to the image endpoint.

Base URL: https://cdn.venz.dev/i/chart.<ext>

  • Supported formats: .svg, .png, .webp, .avif
  • All chart URL parameters (above) are supported, except embed

Additional image parameters

  • w
    • image width in pixels
    • default: 1200, max: 3840
  • h
    • image height in pixels
    • default: 630, max: 2160
  • q
    • image quality (webp/avif only; png is lossless)
    • default: lossless; setting q switches to lossy, range: 1-100
  • pad
    • outer padding in pixels (clamped to half the smaller dimension)
    • default: 0
  • theme
    • color theme
    • default: dark (raster); SVG auto-adapts to viewer/parent unless set
    • values: dark, light, high-contrast

Example:

https://cdn.venz.dev/i/chart.png?type=line&data=3*2*3*4&data=4*3*2*5