Readability Improver
Get specific, actionable suggestions to make your text clearer. Detects long sentences, complex words, passive voice, and more.
Related Tools
How to Use
- Paste Your Text — Enter or paste the text you want to improve. The tool works best with essays, blog posts, emails, reports, and any prose content of at least 2 sentences.
- Review Suggestions — Click Analyze to see your Flesch readability score and a prioritized list of improvements. Each suggestion is color-coded by severity (red = high priority, amber = medium, indigo = low) with the specific issue, fix recommendation, and the affected text.
- Apply and Re-Analyze — Edit your text based on the suggestions, then click Analyze again to see your improved score. Repeat until you reach your target readability level. Copy the improvement report for reference.
About the Readability Improver
The Readability Improver points to the specific sentences and words that make a text harder to read, and says what to do about each one. It reports a Flesch Reading Ease score from the same engine as the Readability Checker, then lists suggestions sorted by priority: sentences to split, passive constructions to rewrite, wordy terms with plainer alternatives, overlong paragraphs, filler adverbs and repetitive sentence openings. It suits anyone revising for a general audience, such as bloggers, support writers, HR and policy teams, or students trimming an essay.
What it checks
| Check | Rule | Priority |
|---|---|---|
| Very long sentence | More than 30 words | High |
| Long sentence | 26 to 30 words | Medium |
| Passive voice | A form of “be” (am, is, are, was, were, be, been, being), optionally one adverb, then a past participle. Common -ed adjectives such as excited, tired or qualified count only when “by” follows within the next few words. One flag per sentence. | Medium |
| Long paragraph | 6 or more sentences (paragraphs are separated by a blank line) | Medium |
| Complex word | Any of 99 listed words with a plainer alternative, such as utilize → use or commence → start, begin; each word is listed once with its count | Low |
| Filler adverbs | 20 words such as very, really, actually, just and quite, flagged when together they exceed 3% of all words | Low |
| Repeated openings | The same first word starts 3 or more sentences, in a text of at least 5 sentences | Low |
Flesch Reading Ease is 206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words), limited to 0–100, with a grade label from Flesch’s scale (for example, 60–69 is 8th–9th grade). You need at least two sentences to run the analysis.
Worked example
Input (5 sentences, 60 words): The new schedule was approved by the board last week. It was designed to facilitate shorter meetings, and it will commence on Monday, although several team leads have said they are really not sure that the additional planning time will be sufficient for the larger projects. It is very simple. It replaces the old rota. Everyone should check their calendar.
- Flesch Reading Ease 66 (8th–9th grade), 12 words per sentence on average
- High: very long sentence (36 words), the second sentence
- Medium: passive voice, “was approved” and “was designed”
- Low: facilitate → help; commence → start, begin; sufficient → enough; additional → more, extra; filler adverbs (2 of 60 words, 3.3%); 3 sentences start with “it”
The score alone looks acceptable because the short sentences pull the average down, but one 36-word sentence holds most of the difficulty. A rewrite that splits it, makes the board the subject and swaps the flagged words (The board approved the new schedule last week. It is meant to keep meetings short, and it starts on Monday. Several team leads doubt that the extra planning time will be enough for larger projects. The schedule replaces the old rota. Everyone should check their calendar.) scores 78 (7th grade) with one suggestion left: “is meant,” a passive you might reasonably keep.
Limitations
- Rules, not understanding. The checks are pattern-based and written for English. They do not know what your sentence means.
- Passive detection has gaps both ways. It misses passives built without “be” (got delayed) and can flag a participle used as a description (the store was closed).
- Complex words are matched without context. “Objective” is flagged even when it is an adjective, and only listed forms are caught: facilitate and facilitates are in the list, facilitated is not.
- Flesch measures only length. It counts words per sentence and syllables per word, so a short sentence full of jargon can still score well.
- Lists and headings need punctuation. Lines without end punctuation may be read as one long sentence; separate paragraphs with a blank line.
Practical tips
- Work from the top of the list. Split a long sentence at its joints (and, although, which) before changing individual words, then click Analyze again.
- Keep a passive when the actor is unknown or does not matter (“The samples were collected overnight”), and rewrite it when hiding the actor makes the sentence vague.
- Treat word swaps as prompts, not orders. A legal or technical term may need to stay exactly as it is.
- Pick a target that fits the audience rather than chasing the highest number; the reading level guide covers how to choose one, and Active vs Passive Voice has examples for the passive flags. For spelling and grammar, use the Grammar & Spell Checker.
Privacy: the rules and the word list are part of the page’s JavaScript, so your text is analyzed in your browser and never uploaded.
Frequently Asked Questions
How is this different from the Readability Checker?
The Readability Checker SCORES your text with 6 formulas (Flesch, Gunning Fog, etc.) and tells you the grade level. This Readability Improver tells you HOW TO FIX specific issues: which sentences to split, which words to simplify, where passive voice exists, and which paragraphs are too long. The Checker diagnoses, the Improver prescribes.
What kinds of issues does it detect?
Six categories: (1) Long sentences over 25 words that should be split, (2) Complex words that have simpler alternatives (e.g., "utilize" to "use"), (3) Passive voice constructions that should be rewritten in active voice, (4) Long paragraphs with 6+ sentences that should be broken up, (5) Overuse of filler adverbs like "very", "really", "actually", and (6) Repetitive sentence openings where multiple sentences start with the same word.
How many complex words does it know?
The tool contains a database of about 100 complex words and word forms mapped to simpler alternatives. Examples: "utilize" to "use", "facilitate" to "help", "approximately" to "about", "demonstrate" to "show", "methodology" to "method", "necessitate" to "need", "remuneration" to "pay". Each suggestion shows the exact simpler word to use.
What is a good Flesch readability score?
It depends on the audience. On Flesch's scale, 90-100 corresponds to 5th grade, 70-79 to 7th grade, 60-69 to 8th-9th grade and 30-49 to college level. Many writers for a general web audience aim for 60 or higher, while academic and technical writing often scores lower because it needs longer, specialized words. Treat the number as a guide rather than a pass mark. The tool shows your current score with a visual bar and grade level, so you can track improvement as you edit. The score is calculated with the same engine as our Readability Checker, so both tools report the same Flesch value for the same text.
Does it use AI to generate suggestions?
No. The tool uses rule-based analysis: patterns that look for a form of "to be" followed by a past participle to spot passive voice (common adjectives such as "excited" only count when followed by "by"), a curated word complexity database for simplification suggestions, word counting for sentence length, and string matching for repeated sentence starts. No AI, no API, no machine learning. The rules are predictable and private, but like any rule-based checker they can occasionally miss an issue or flag a sentence that is fine as written.
Is my text sent to a server?
No. All analysis runs in your browser using JavaScript. Your text is never transmitted to any server. The complexity database, passive voice patterns, and readability formulas are all embedded in the client-side code.