Distributions
Make a histogram from the numbers you already have
Type the bins and their counts in a sentence and the chart is drawn for you — bars touching, on your own colours, with nothing to drag.
Straight answer first: there is no canvas here and nothing to drag. You describe the histogram in a sentence, or bring a file you already have, and it is drawn for you — on your own brand. If you want to place every shape by hand, draw.io is free and better at that. If you have the numbers already, read on.
Your file is read once to draw the diagram and is never stored. Converting turns on anonymous usage analytics — we never send your file, its contents or your brand URL. What we measure
Paste any URL. We read its palette and type and redraw the diagram in them — about a second, no account.
Colour
Found on your site
Any colour
Found on your site
Any colour
Found on your site
Any colour
Found on your site
Any colour
Found on your site
Any colour
Found on your site
Any colour
Or set it yourself
Every choice here redraws the diagram. Nothing is saved.
Language
The layout mirrors and the Arabic type scale applies, free. Translating the labels needs an account on a paid plan.
Skin
Type
Motion
Line
Depth
A bar in relief still measures correctly — the top face sits on the bar's end. A pie does not: squashing the circle makes the near slices read larger than the far ones at the same share.
Show
Font
Your diagram
Selected shape
Icon
Fill
Border
Label
Pages in this file
Try it with this
Paste this into the description box above — it is the example that produces the worked histogram in the gallery.
How many diagrams had 1-5 nodes: 14. 6-10: nine. 11-20: six. 21-40: three. Over 40: two.
A histogram is not a bar chart
The difference is the gap, and it is not decoration. A bar chart compares separate categories — sales by region, headcount by team — so its bars stand apart. A histogram shows one continuous measure cut into ranges: ages, response times, file sizes. Its bars touch, because a gap would claim the data stops between bins, which is exactly what a continuous variable does not do. Draw a distribution with gaps and you have told the reader something untrue about the data.
What comes out
Drawn by this pipeline, not picked to flatter it.
What you get out
SVG and PNG on any plan. PowerPoint with real editable shapes and live text — not a picture of a chart you cannot fix a typo in — plus our own .dgm format, which is plain text you can edit and re-import. Every conversion returns a ledger saying what changed, because you know your own diagram and would notice anyway.
Questions people actually ask
- Do I need the raw data?
- No — the counts per bin are enough. If you already have bins and frequencies, that is the whole input. Raw observations still need binning somewhere first; a spreadsheet does that part better than we would.
- How many bins should I use?
- Enough to show the shape and few enough to read — usually five to fifteen. That is a decision about your data rather than about the chart, so it is left to you rather than guessed.
- Can I put it in a slide?
- Yes. Export to PowerPoint and the bars arrive as real editable shapes with live text, not a picture of a chart you cannot correct.
Other diagram types
flowchart maker swimlane diagram process flow diagram pie chart maker venn diagram maker gantt chart maker org chart maker er diagram tool
All 53 types have a worked example in the gallery.