Word Counter — Free Online Tool on Toolpile

Pricing

Every tool you need, one place

PDF, Image, AI, Dev, and Business tools — free, private, no signup.

PDF

20

Image & Media

17

AI & Prompts

21

Text

13

Developer

26

Converters

6

SEO & Marketing

8

Generators

12

Productivity

7

Calculators

4

Business

9

Privacy & Security

5

Design

4

Content

5

Education

4

About Word Counter

A word counter looks trivial until you realise how many different things 'a word' can mean — and how many real-world limits (Twitter, LinkedIn, college essays, SEO meta descriptions, SMS) care about one specific count rather than a generic one. This tool handles the counts; what follows is how to use them without getting caught out.

Words, characters, and why they disagree

The word count you see in Microsoft Word, Google Docs, and this tool all agree within ±1 for typical prose — they split on whitespace and count the pieces. Where they diverge is punctuation-attached tokens ("don't" is one word; "state-of-the-art" is one in Word but four in some academic counters), URLs (Word counts them as one word, most web counters split on the dots), and numbers with units ("5 km" = 2 words; "5km" = 1).

Character counts are more stable but have their own trap: "characters with spaces" vs "characters without spaces." Twitter's 280-character limit counts spaces and emojis (most emojis are 2 code units). LinkedIn's 3,000-character post limit counts them too. SMS messages count bytes, not characters — a single emoji can consume 2-4 bytes of the 160-byte limit depending on encoding. When a platform says "280 characters," always test a real-world sentence, not just your count.

Reading time is derived from word count using an assumed words-per-minute (WPM) rate. Silent reading averages: 200-250 WPM for adult English readers, 100-150 for children, 400-700 for skimming, 600-1,500 for speed-readers. This tool defaults to 225 WPM, the number used by Medium and most blog platforms. Technical or academic text realistically runs 30-40% slower because of paragraph-level re-reading — if your content is dense, manually round the estimate up.

Platform limits worth memorising

Most writing gets clipped somewhere. Keeping a mental model of the common cut-offs saves a lot of re-editing:

  • Twitter/X post: 280 characters (premium: 25,000). Counts spaces; URLs count as 23 chars regardless of length.
  • LinkedIn post: 3,000 characters. Headlines: 220. Bio 'About': 2,600.
  • Instagram caption: 2,200 characters; only first 125 show before 'more'.
  • SEO meta description: ~155-160 characters before Google truncates. Meta title: ~60.
  • SMS: 160 characters GSM-7, 70 if any Unicode char forces UCS-2 encoding — one emoji halves your limit.
  • College essay (Common App): 650 words hard max. UCAS personal statement: 4,000 characters or 47 lines (UK).
  • Academic abstract: 150-300 words depending on journal; check submission guidelines.
  • Email subject line: best performance 40-60 characters; inbox preview shows ~35 on mobile.
How to use this counter
  1. Paste or type your text into the editor. The counts update live — no submit button needed.
  2. Check words, characters (with and without spaces), sentences, paragraphs, and estimated reading time side-by-side.
  3. For platform-specific checks, watch the relevant count: tweets = characters including spaces; blog reading-time = words ÷ 225.
  4. For essay/academic work, use the word count. The sentence count is a rough readability check — if average words-per-sentence exceeds 25, your paragraphs are probably dense enough to lose readers.
  5. Everything runs client-side; nothing is transmitted, stored, or logged. You can paste a draft of a confidential document without it leaving your browser.
FAQ

Does the counter match Microsoft Word exactly?

On plain prose: within ±1 word. On mixed content with URLs, tables, footnotes, or tracked changes, counts can diverge by 5-10%. For official submissions that specify 'as counted by Microsoft Word,' open the document in Word and use Tools → Word Count — no in-browser tool can exactly replicate Word's tokeniser because Word counts things like comments and headers optionally.

Why is my Twitter post rejected even though it's under 280 characters here?

Twitter counts URLs as a fixed 23 characters regardless of their real length via the t.co wrapper, and some emoji sequences (like flags or skin-tone modifiers) count as multiple characters to Twitter even though they render as one glyph. For mission-critical tweets, paste into Twitter's own composer and trust its counter over any third-party tool.

Is 'don't' one word or two?

One. Every major counter treats contractions as single words by splitting on whitespace, not apostrophes. This also means 'rock'n'roll' counts as 1 word, which feels wrong but matches the Word/Docs convention.

How accurate is the reading time?

Accurate to within ~20% for general audiences at the default 225 WPM. Technical docs, contracts, and academic papers read 30-40% slower; light fiction and news articles read 10-20% faster. For a platform like Medium, the 225 WPM figure is what they publish, so your estimate will match theirs.

Does it count Chinese, Japanese, or Arabic correctly?

Character counts yes, word counts partially. Whitespace-free languages (Chinese, Japanese) have no reliable word-count in a language-agnostic tool — the 'words' figure will reflect visual whitespace, which usually approximates sentence count instead. Use the character count for those languages; it's what publishers expect (e.g., a Chinese academic paper of '5,000 characters' is the real unit).

Is there a way to count without counting spaces?

Yes — the tool shows both 'characters including spaces' and 'characters excluding spaces.' Use the exclusive count for fields that specify 'no-space character limits' (rare but some academic style guides and coding interviews do). Use the inclusive count for everything social-media, SEO, and SMS related.

Related tools

Related tools

See all text tools →
Looking for something else? Browse all 155 tools.