Compare two drafts. Pick the cleaner one.

Last updated

Draft A
0 words · 0 sentences
Draft B
0 words · 0 sentences
Paste two texts above. The clearer one wins.

Score-by-score

Lower grade = easier; higher Flesch = easier
MetricDraft ADraft BWinner
Flesch Reading Ease———
Flesch–Kincaid Grade———
Gunning Fog———
SMOG———
Coleman–Liau———
Automated Readability Index———
Consensus grade———

When to use the compare tool

The compare view is built for one job: deciding whether an edit actually made writing simpler. You finish a revision, you think it reads better, but you can't tell whether that's the rush of having edited it or a real improvement. Paste both, and the formulas vote.

Useful for:

  • A/B testing two openers on a landing page or email subject line
  • Comparing your draft against a competitor's published article
  • Reviewing AI-generated text against your hand-edited version
  • Translating from technical English to a plain-language version
  • Checking that "shortening" actually lowered the grade level — sometimes it doesn't

If you only want one number, watch the consensus grade at the bottom: it's the mean of all six grade-equivalent scores, so no single formula decides the result. Lower wins.

How the winner is picked

Each draft gets the same six scores you see on the readability checker: Flesch Reading Ease, Flesch–Kincaid, Gunning Fog, SMOG, Coleman–Liau and the Automated Readability Index. Every row in the table names the easier draft for that formula. For Flesch Reading Ease, the higher score wins; for the other five, the lower grade wins.

The overall verdict uses the consensus grade. Flesch Reading Ease is converted to a grade first, then all six are averaged. A gap of less than 0.1 grade is reported as a tie.

Example: a corporate paragraph and a plain rewrite

These two drafts were written for this page and say roughly the same thing. The scores come from this calculator, so you can paste them above and get the same numbers.

Draft A: “Our quarterly evaluation identified significant opportunities to improve operational efficiency. Management will prioritize initiatives that reduce administrative complexity. We expect measurable results by the end of the year.”

Draft B: “We checked how the team works each quarter. We found ways to cut busywork. Next, we will fix the steps that slow people down. You should see results by the end of the year.”

FormulaDraft ADraft BEasier
Flesch Reading Ease0.0100.0Draft B
Flesch–Kincaid Grade19.21.3Draft B
Gunning Fog22.34.6Draft B
SMOG14.45.7Draft B
Coleman–Liau22.43.9Draft B
Automated Readability Index17.02.0Draft B
Consensus grade18.63.7Draft B

Draft A has short sentences, but nearly half its words have three or more syllables. Draft B is longer by six words and still reads about 15 grades easier, because its words are short. The Gunning Fog page works through the same pair step by step.

Tips for a fair comparison

  • Compare the same content. Two drafts on different topics tell you about the topics, not the writing.
  • Keep the lengths close. A 40-word sample and a 400-word sample are not scored on equal terms.
  • Paste plain text. Headings, bullet fragments and code can split or merge sentences and skew the counts.
  • Read the per-formula rows, not just the verdict. If the formulas disagree, look at sentence length and word length separately. The readability formulas guide explains what each one measures.

Frequently asked questions

How does the compare tool pick a winner?

It compares the consensus grade of each draft, the average of six formulas, with Flesch Reading Ease converted to a grade first. The draft with the lower grade wins. A gap under 0.1 grade is called a tie.

Is my text sent anywhere?

No. Both drafts are scored in your browser. Neither is uploaded or stored.

Can a shorter draft score worse?

Yes. Cutting words does not always lower the grade. If you remove short words and keep long ones, syllables per word go up, and so can the grade. Comparing shows whether an edit helped.

How long should each draft be?

Use at least a few sentences. Very short samples swing a lot, and SMOG was designed for 30 sentences. For a fair test, compare passages of similar length that cover the same content.