Word & Character Counter
Text Tools
Count words, characters, sentences, paragraphs, and lines instantly. See reading time, speaking time, and top keywords. Updates as you type.
Platform character limits
Counts raw Unicode code points. X URL weighting and SMS encoding rules are different.
Adjust reading / speaking speed
Estimated AI tokens
Rough estimate based on ~4 characters per token (the widely-used rule of thumb for GPT-4, Claude, Gemini, and similar models). Code, numbers, and non-English text tend toward the high end; plain English prose lands near the typical figure.
How to use this tool
- Paste or type your text. Paste your document into the text area. Results update automatically as you type.
- Read the stats. Word count, character count, platform-limit indicators, sentences, paragraphs, lines, reading time, and speaking time appear as you type.
- Adjust reading speed (optional). If your audience reads faster or slower than average, change the words-per-minute value to get a more accurate time estimate.
How word count varies by counting method
"Word count" sounds like a single number, but different tools and style guides count differently. This counter uses whitespace-separated tokens that contain at least one letter or digit. That means "hello," counts as one word (the comma is attached but stripped for counting), and "---" counts as zero words (no alphanumeric characters).
Microsoft Word, Google Docs, and most publishing tools use similar rules, but they differ on edge cases such as hyphenated compounds, contractions ("don't"), and words made of numbers only ("2024"). Here, "self-aware" and "don't" each count as one word, and numeric tokens like "42" count as words. Han, Hiragana, Katakana, and Hangul characters count individually; this simple convention is not full linguistic word segmentation.
Reading speed averages
The default reading speed of 200 words per minute is a frequently cited average for adult silent reading of ordinary prose. The range across adults is wide: slow readers manage around 100–150 wpm, average adults 200–250 wpm, and speed readers above 300 wpm, though comprehension tends to drop at the high end. Academic reading of dense material is slower, often 100–150 wpm.
Speaking time uses 130 wpm, close to a natural conversational pace. Presenters typically speak at 120–150 wpm; auctioneers and fast talkers push above 250 wpm. For scripted speeches and presentations, 125–150 wpm gives enough time for emphasis and breathing without rushing.
Use the speed controls below the stats to calibrate for your audience. A technical talk for experts can move faster than a keynote for a general audience.
Character count use cases
The live indicators above compare your text's Unicode code-point count with common raw character ceilings. They are useful for drafting, but platform rules can differ: X applies special weighting to URLs, and SMS segment limits depend on the message encoding. Use a Social Character Counter for X's weighted post length and consult your carrier for SMS details.
| Platform / context | Limit |
|---|---|
| Twitter / X post | 280 characters |
| SMS (GSM-7 encoding) | 160 characters per segment |
| SMS (Unicode / emoji) | 70 characters per segment |
| Meta description (SEO) | ~155 characters |
| SEO title tag | ~60 characters |
| App store title (Apple / Google) | 30 characters |
| LinkedIn post | 3,000 characters |
Readability scores
This counter reports word and sentence statistics but does not calculate a readability score. For Flesch Reading Ease, Flesch–Kincaid grade level, and four other readability formulas, use the Reading Level Estimator.
Academic word count guidelines
| Document type | Typical word count |
|---|---|
| Abstract | 150–250 words |
| Essay | 500–2,000 words |
| Research paper | 3,000–8,000 words |
| Thesis chapter | 8,000–12,000 words |
| Master’s thesis | 15,000–25,000 words |
| PhD dissertation | 60,000–100,000 words |
Frequently asked questions
How is word count calculated?
Latin and numeric words are whitespace-separated tokens that contain at least one letter or digit, so hyphenated compounds like "self-aware" count as one word. Han, Hiragana, Katakana, and Hangul characters are counted individually. Punctuation-only tokens are ignored.
Does the counter work with emoji and non-English text?
Yes. Characters are counted using Unicode code points, so emoji like 😀 count as one character rather than two (which is what JavaScript's default .length gives). Han, Hiragana, Katakana, and Hangul characters count individually as words; this is a simple character-based convention, not linguistic word segmentation.
What reading speed is used for the time estimate?
The default reading speed is 200 words per minute (wpm), which is a commonly cited average for adult silent reading. Speaking time uses 130 wpm, close to the average conversational speaking pace. Both can be adjusted in the settings below the results.
What are the top keywords?
The 10 most frequent terms after common English stopwords are removed. Each row shows its count and its share of all counted words; the percentage is a rough density signal, not an SEO target or a substitute for context.
Related tools
- Reading Level EstimatorCheck Flesch-Kincaid, Fog, Coleman-Liau, SMOG, and ARI readability scores for your text.
- Keyword DensityReview keyword frequency and density, including phrase-level analysis for SEO drafts.
- Social Character CounterCheck platform-specific social post limits, including X URL weighting and other networks.
- LLM Token CounterEstimate model-specific LLM token counts with a tokenizer instead of a rough character ratio.