Character counter. Live, with thresholds.
Type or paste text. We count characters, words, sentences, paragraphs, lines, and reading time live. Plus colored thresholds for the formats that matter — tweet length, SEO title, meta description, ad copy.
Type or paste in the box. Counts update live as you type — characters, words, sentences, paragraphs, lines, reading time. Plus colored threshold badges for the formats that matter: tweet (280), Bluesky (300), SEO title (60), meta description (155), Google Ads. Browser-only — your text never leaves the page.
Privacy: text is processed in your browser. Nothing is sent or logged.
Six metrics. Different jobs.
Characters is the byte-true count for any platform that imposes a hard cap. Tweets, SMS messages, push notifications, Google Ads creatives — all enforce by character. We count by Unicode code point, not byte, so emoji and combining-character sequences count as the visual character set the platforms enforce against.
Words is the writer's pacing metric. We split on whitespace; hyphenated terms ('state-of-the-art') count as one, contractions ('don't') count as one, URLs count as one. Useful for estimating effort, comparing drafts, and writing to a target length. For SEO, word count is a soft signal — comprehensive content tends to rank better than thin content, but no specific count guarantees ranking.
Sentences and paragraphs are the readability signal. Long sentences (over 25 words) and long paragraphs (over 80 words) hurt scan-reading per NNGroup's web reading research. If the sentence count is low and the word count is high, sentences are too long. Pair with our Readability Checker for the Flesch-Kincaid grade.
Reading time is the audience expectation. We use 238 words per minute — the average adult silent-reading speed for medium-difficulty English per Brysbaert (2019), a meta-analysis of 100+ reading-rate studies. The figure is a rough guide: technical content reads slower (150-200 wpm), narrative reads faster (250-300 wpm). Show it as "12 min read" on long-form content; readers self-select.
Top words is the self-edit crutch-check. We exclude common stopwords (the, a, is, of, and, …) so the list reflects what the draft is actually about. If your top non-stopword is the topic itself, you're on-message. If it's "just" or "really" or "actually," you have a tic to clean up. Six rounds of this self-edit and the prose tightens dramatically.
Five jobs this tool covers.
Job 1: Write to a hard limit. Drafting a tweet, an ad headline, a meta description, an SMS — the threshold panel gives you live red-amber-green feedback as you type. No more "post too long" errors after the fact. Especially useful for ad-copy work where 10 different formats each have their own caps.
Job 2: SEO on-page hygiene. Title tag ≤ 60 chars, meta description ≤ 155 chars are the live thresholds Google enforces with truncation. The badge flips amber as you cross 80% so you have time to tighten before going over. Pair with our Title Tag Checker and Meta Description Checker for full SERP-snippet validation including pixel-width constraints.
Job 3: Long-form drafting. Show the live word count as you write a blog post or report; aim for the band you've decided is right (1,200-1,800 for a feature piece, 600-900 for a quick take). Reading time on the published article tells the audience what they're committing to before they start.
Job 4: Self-edit pass. The top-words list is a free draft-tightening tool. Crutch words ('just,' 'really,' 'actually,' 'very,' 'kind of') near the top of the list mean a tightening pass will improve the prose without changing the substance. Hemingway's app does the same job in a different format; this is the no-signup version.
Job 5: Reviewing AI output. When ChatGPT or Claude returns prose, the top-words list will show you which adjectives the model leaned on. AI-generated content tends to overuse a small vocabulary of abstract words. Spot them in the top-8 and rewrite — the difference between AI-stamped and edited-AI is the rewrite of those crutch words.
Six questions users ask.
How are words counted?
We split on whitespace and count any non-empty token as a word. This matches the behavior of Google Docs, Microsoft Word, and the WordPress editor — hyphenated terms ('state-of-the-art') count as one word, contractions ('don't') count as one, URLs count as one. The total is exact for body copy and short for content with many compound nouns; that's the standard convention. For a strict typographic count, use the character total instead.
What are the threshold badges for?
Five formats with hard limits matter at the top of most content tasks. Tweet (X / Twitter): 280 chars. Bluesky post: 300 chars. SEO title tag: 60 chars (Google truncates with ellipsis past that). Meta description: 155 chars (Google truncates rendered descriptions past that on most queries). Google Ads headline: 30 chars per slot, description 90. We render each as a small badge that flips green / amber / red as you cross the threshold so you can write to-format without switching tabs.
How is reading time calculated?
Words divided by 238 — the average adult silent reading speed for medium-difficulty English text per Brysbaert (2019), the most-cited reading-rate meta-analysis. Below 1 minute we show seconds. Use it as a rough estimate; actual reading speed varies 100-450 wpm depending on familiarity with the topic, age, and reading purpose. For long-form content, the reading-time figure is most useful as a relative comparison between drafts, not an absolute promise.
Does it count Unicode and emoji correctly?
Yes — we use the JavaScript spread operator on the string, which iterates by Unicode code points rather than UTF-16 code units. Emoji compounds (👨👩👧👦), regional indicators (🇮🇳), and combining-character sequences count as the visual character set, not the underlying byte count. This matches what platforms like Twitter and Bluesky use to enforce their character limits.
Is the text I paste sent anywhere?
No. Everything happens in your browser. Your text is never sent to Digital Heroes servers, never logged, never analyzed by us. The page is static HTML; the only network request is the initial page load. Safe for sensitive drafts, NDA'd copy, or pre-publication content.
Why is the most-common-words list useful?
It's a quick crutch-word check. If your top three non-stopwords are all the same theme, your draft is on-topic. If your most-frequent non-stopword is something like 'just' or 'really' or 'actually,' you have a writing tic. We exclude common stopwords (the, a, is, of, and, …) so the list reflects content words, not grammar. Useful as a self-edit pass before publishing.