Free Burstiness Checker Chart Your Sentence-Length Spread
Burstiness measures how widely sentence lengths swing across a passage, not how long the average sentence happens to be. Two passages can share an identical mean length and read nothing alike: one holds every sentence close to that mean, the other alternates short and long. This checker plots every sentence as a point, computes the spread as a coefficient of variation, and grades that spread against what edited prose typically shows. Rows of same-length sentences are one of the fastest ways a passage reads mechanical; this page puts the actual distribution on screen.
Charts the length of every sentence and scores the variety of your rhythm against edited human prose.
0 / 1,000 words
84,000+
International Researchers
Average Rating
Trusted by researchers publishing in
How the burstiness checker reads a passage
Load in the passage to chart
One paragraph gets a reading started; a full section shows the truest distribution.
Study the point-by-point chart
Each point is one sentence, plotted by length. A tight cluster near the line signals a flat, machine-like structure; a wide scatter signals variation. Exact word counts sit below the chart.
Revise where the spread is thin
Shorten each paragraph's central claim, merge thin neighbours, split overloaded chains, then re-chart and watch the distribution widen.
What the burstiness score is built to show
The full distribution, charted
Rather than a single number, the chart plots every sentence as its own point, so a cluster near the mean or a scatter across the range is visible at a glance.
Judged relative to your own mean
The spread is computed against your passage's own average sentence length, so a dense theory section and a plain summary are held to the same standard regardless of how long their sentences run.
Bands set for research prose
The reference bands come from academic and professional writing, with room for methods sections and procedural lists that are legitimately uniform.
A distribution statistic, nothing more
The page reports sentence-length data. No claim about who wrote the passage rides along with the number.
Sentence-length distribution and why it varies
Read a well-edited paragraph aloud and the sentence lengths visibly disagree with each other: a long one lays out a case, a short one closes it. That disagreement is a distribution, not a coincidence, and stylometrists have a name for how widely a set of sentence lengths spreads around its own mean: burstiness. It sits one level below perplexity, which tracks word-by-word predictability; burstiness tracks the shape of the sentences themselves, independent of which words fill them.
The measurement itself is a coefficient of variation: take the standard deviation of words-per-sentence and divide it by the mean. Two passages can carry the identical average sentence length and score at opposite ends of this scale, because the average tells you nothing about how the individual sentences are arranged around it. A model generating the statistically likely next sentence gravitates toward the middle of its own range for the same reason it gravitates toward common words, so raw AI output frequently produces a run of sentences clustered within a narrow band of lengths.
Widening that band comes from editing for emphasis. Chasing the number on its own produces sentences chopped at random, which score well and read badly. Locate the sentence in each paragraph that carries the actual claim and shorten it. Merge two thin sentences that share a subject. Split any sentence stacking three clauses on commas and conjunctions. The distribution widens as a side effect of edits that also make the paragraph easier to follow.
Sentence length is one axis of a passage's shape. Word-level predictability is a separate axis, measured by the perplexity checker, and the specific vocabulary a model overuses is covered by the AI word cleaner. Rebuilding the distribution across an entire manuscript, rather than charting it, is what the TextPulse humanizer does.
Reading the shape behind the score
Each sentence becomes one point on the line, plotted by its word count. The distribution the line traces carries more information than the single number above it.
Every point sits within a few words of its neighbours, the signature of a passage whose sentence lengths barely move off the mean. Raw output from a language model often draws a line exactly like this one, and a reader feels the sameness before naming it.
Long sentences carrying an explanation sit beside short ones delivering a verdict, and the line swings between them. This is the distribution shape of prose someone has read aloud and cut where the breath naturally stops.
A useful spread has a reason behind every point on the chart: each long point doing explanatory work, each short point landing a claim that earns the emphasis. A line that never leaves the average belongs to prose nobody has read back aloud.
Who uses the burstiness checker
Coursework and essay writers
Notice why a paragraph reads flat before a marker circles it, and pick up the trick of following a long sentence with a short one.
Academics preparing submissions
Check whether a section drafted quickly, or with AI assistance, has gone rhythmically flat before a reviewer feels it first.
Brand and content teams
A uniform rhythm reads as templated to a person and to a ranking system alike. Chart a draft before it carries a client's name.
Writers drafting in a second language
Sentence length often flattens out when writing in a second language, without the writer noticing. The chart makes the pattern visible immediately.
One flat rhythm, made visible.
Rebuilding it takes a full pass.
The TextPulse humanizer rebuilds sentence rhythm, word choice and phrasing across an entire draft, with every change shown as a tracked edit and citations left exactly where they were.