CountWise Word Counter
Exact word, character, and sentence counts with essay and social-post limit checks.
CountWise Word Counter counts words, characters (with and without spaces), sentences, paragraphs, and unique words with documented, deterministic rules, then estimates reading and speaking time, English readability, and keyword density. It can leave out citations, quotations, headings, and the reference list for essay limits, and checks text against X, Threads, Bluesky, LinkedIn, Instagram, SEO title and description, and SMS limits, showing exactly where to cut.
What it does
- Word, character (with and without spaces), letter, sentence, paragraph, line, and unique-word counts
- Essay mode: leave out APA/MLA in-text citations, quoted passages, headings, and the reference list
- Reading time (238 wpm), speaking time (150 wpm), Flesch reading ease and Flesch-Kincaid grade for English
- Top keywords with density, the longest sentence, and warnings for sentences over 25 and 40 words
- X/Twitter weighted counting (links 23, emoji and CJK 2), Threads, Bluesky graphemes, LinkedIn, Instagram
- SEO title and meta description lengths, SMS GSM-7 and Unicode segment counts, and where to cut
- Word segmentation for Japanese, Chinese, and Thai, plus CJK character counts
Try asking ChatGPT
“Count the words in my essay excluding citations and the reference list”
“Does this post fit X's 280 limit and LinkedIn's 3,000 limit?”
“How long will this speech take to read aloud?”
Limitations
- Up to 100,000 characters per request; split longer documents
- Citation, heading, and reference-list detection follows common APA/MLA patterns and may miss unusual formats
- Readability scores are English-only estimates; word splitting for Japanese, Chinese, and Thai uses dictionary segmentation, so totals can differ from other tools
- Platform limits change over time; checks use the published limits as of October 2026. Files (PDF, Word) cannot be read directly; paste the text
Connect
Install CountWise Word Counter from the ChatGPT plugin directory. Developers can connect the MCP endpoint directly: https://word-counter.zovo.one/mcp