Box Plot Maker & Box and Whisker Plot Generator
Paste your numbers or drop a CSV or Excel file — the quartiles, medians, whiskers and outliers are computed exactly, and the AI draws a publication-ready box plot in your target journal’s conventions. Refine it in chat and export when it’s right.
Box plot styles
One real render per journal preset, each from a different dataset — click a card to apply the style, or zoom for the full image.
What is a box plot?
A box plot (box and whisker plot) summarizes a numeric dataset with five values: the minimum, first quartile, median, third quartile and maximum. The box spans the middle half of the data from Q1 to Q3, the line inside it marks the median, the whiskers reach the most extreme points within 1.5 × IQR of the box, and anything beyond them is drawn as an individual outlier point.
About this box plot maker
- Paste values straight into the prompt or attach up to three CSV, TSV or Excel files — one column per group, or label/value rows.
- Quartiles, medians, whiskers and outliers are computed exactly — linear-interpolation quartiles with Tukey 1.5 × IQR fences — before the AI draws a single line.
- Journal style presets format the same statistics to Nature, Science, Cell Press, clinical-grayscale or dark-slide conventions.
- Every follow-up in the workspace chat recomputes the statistics from your data, so edits never drift from the numbers.
- Finished figures export as PNG, JPG, SVG or PPTX through the canvas editor.
How outliers are defined
A point is an outlier when it falls more than 1.5 interquartile ranges outside the box. The whiskers end at the most extreme data points still inside those fences — not at the fences themselves.
IQR = Q3 − Q1 fences: Q1 − 1.5 × IQR and Q3 + 1.5 × IQR
In words
The interquartile range is the third quartile minus the first. Any value below the lower fence or above the upper fence is plotted as its own point instead of being included in a whisker.
The 1.5 × IQR rule is Tukey’s original convention and the default in R, Python and most statistics software.
How to make a box plot
Add your data
Paste one group per line — Control: 4.1, 5.2, 4.8… — or drop a CSV or Excel export, one column per group or label/value rows. The exact five-number summary of every group appears instantly, outliers included.
Describe the figure
Say what the axes mean and how it should read — y-axis “Response time (s)”, horizontal, grayscale for print. Pick a journal style or leave the preset on auto.
Generate and refine
The AI draws the plot from the computed statistics, never from guesswork. Keep editing in chat — rename a group, flip it horizontal, switch journal styles — then export as PNG, SVG or PPTX.
Parts of a box plot
Every element of a box and whisker plot answers one specific question about your data.
| Part | What it shows | How it is computed |
|---|---|---|
| Box | The middle 50% of the data | Spans Q1 to Q3 — the interquartile range |
| Median line | The middle value | Half the observations sit on each side |
| Lower whisker | Data reach below the box | To the smallest point at or above Q1 − 1.5 × IQR |
| Upper whisker | Data reach above the box | To the largest point at or below Q3 + 1.5 × IQR |
| Whisker caps | Where the whiskers end | Short bars at the last data point, never at the fence itself |
| Outliers | Unusually extreme values | Individual points beyond the 1.5 × IQR fences |
| Mean marker | The arithmetic mean (optional) | Drawn on request, usually as an open diamond |
| Value axis | The measurement scale | Regular tick steps covering every group, whiskers and outliers |
Quartiles use linear interpolation (R type 7) — the same method as NumPy, Excel and Google Sheets, so the numbers here match your spreadsheet’s.
Box plot vs histogram vs violin plot
Three ways to show a distribution — which one fits depends on what the reader should compare.
| Question | Box plot | Histogram | Violin plot |
|---|---|---|---|
| What it shows | Five-number summary plus outliers | The full shape of one distribution | A smoothed density with a summary inside |
| Best for | Comparing several groups side by side | One variable’s distribution in detail | Comparing distribution shapes of larger samples |
| Works from | As few as ~5 points per group | Enough points to fill meaningful bins | Enough points to estimate a density |
| Outliers | Marked as individual points | Hidden inside the end bins | Usually not marked |
| Bimodality | Invisible — the box can’t show two peaks | Visible as two peaks | Visible in the density outline |
If the story is “group A runs higher than group B”, a box plot says it fastest. Reach for a histogram or violin plot when the shape of the distribution is itself the story.
Where box plots are used
Any time several groups of measurements need comparing at a glance, a box plot is the standard answer.
Treatment comparisons
Dose–response experiments, drug vs vehicle, pre vs post — the side-by-side boxes make shifts in median and spread visible instantly, with outlier animals or samples flagged honestly.
Quality control
Batch-to-batch variation, instrument drift, inter-operator spread — a box per batch shows both the level and the consistency, and outlier runs stand out as points.
Teaching statistics
The five-number summary, quartiles and the 1.5 × IQR outlier rule all live in one figure — which is why box plots anchor descriptive-statistics units from middle school through university.
Survey and A/B results
Response times, scores and ratings across cohorts or variants — box plots compare a dozen segments in one figure where a table of means would hide the spread.
Box plot questions, answered
How do you read a box plot?
Read the median line first, then the box, then the whiskers. The line inside the box is the median — half the data sits on each side. The box covers the middle 50% (Q1 to Q3), so a taller box means more spread. The whiskers show the range of typical values, and any separate points beyond them are outliers worth a second look.
What do the whiskers on a box plot show?
The whiskers span the data outside the box that still counts as typical — from each box edge to the most extreme observation within 1.5 interquartile ranges. They end on a real data point, never on the fence value itself, which is why the two whiskers are usually different lengths.
How are the quartiles calculated?
This tool uses linear interpolation between order statistics — the method NumPy, Excel and Google Sheets use by default (R’s type 7). Textbooks sometimes teach a different rule (splitting the halves around the median), which is why two tools can give slightly different Q1 and Q3 for the same data. Whichever the method, the median itself always agrees.
Can I use this as a box plot calculator?
Yes — before anything is drawn, the tool works as a box and whisker plot calculator: paste data and the exact n, min, Q1, median, Q3, max and outliers of every group appear instantly. The same table doubles as a five number summary calculator and an interquartile range calculator — IQR is Q3 − Q1, and the 1.5 × IQR outlier fences are applied automatically. Checking the statistics costs nothing; generation is only spent when you draw the figure.
What counts as an outlier in a box plot?
Any value more than 1.5 × IQR below Q1 or above Q3. This is Tukey’s convention, and it flags roughly 0.7% of normally-distributed data — so an occasional outlier point is expected, not automatically an error. Outliers are drawn as individual points so the reader can count them.
Can I make a box plot from an Excel or CSV file?
Yes — attach up to three .xlsx, .csv or .tsv files with the composer’s upload control, or paste the values straight into the prompt. One column per group with a header row works, and so does the long format of a label column plus a value column. The exact statistics of every group appear immediately, before any generation is spent.
How many groups can one box plot compare?
Up to 12 groups on one axis. Beyond that the boxes get too narrow to read and reviewers will ask you to split the figure anyway. If your file has more columns, the panel keeps the first 12 and tells you what it skipped.
Can I make a horizontal box plot?
Yes — just say “horizontal” in the description. Horizontal boxes are the better choice when group names are long, and standard in clinical journals. Everything else — statistics, styles, editing — works identically in both orientations.
Does the AI change or invent my numbers?
No. The quartiles, whiskers and outliers are computed exactly from your data by code, and the AI is instructed to draw precisely those values. Every regeneration recomputes from the original data, so iterating on the style never drifts the statistics.
Is a box plot the same as a box and whisker plot?
Yes — the two names describe the same chart. “Box and whisker plot” (or box-and-whisker diagram) is the longer teaching name; “box plot” or “boxplot” is what statistics software and journals call it. Both were popularized by John Tukey in 1977. The tools inherit every combination of the names — box plot maker, boxplot generator, box and whisker plot creator — and they all draw this same chart.
Can I show the mean on a box plot?
Yes — ask for a mean marker and each box gets one, usually an open diamond. The box plot itself is built around the median because it resists outliers; showing both is standard when readers will want to connect the figure to reported means.
What export formats are available?
PNG and JPG directly, and SVG or PPTX through the canvas editor, which splits the figure into editable elements first. That covers manuscript submission (vector), slide decks (PPTX) and quick sharing (PNG).
Is this box plot maker free?
Yes to start — every new account gets 100 free credits and no card is required to sign up. That covers generating and exporting real figures before you decide whether to buy credits or subscribe. An account is needed because each figure is saved to a workspace you can return to and edit.
Sources
- NIST/SEMATECH e-Handbook of Statistical Methods — Box Plot — National Institute of Standards and Technology
- Visualizing samples with box plots (Points of Significance) — Nature Methods
- quantile — R documentation of the nine quantile types — The R Project
Keep exploring
Our other diagram tools, and the guides behind them.
Draw your box plot
From pasted numbers to a publication-ready figure in about a minute — statistics computed exactly, styling handled by the AI.
Start drawing