When to use it
You're drafting a Korean self-intro essay (자소서) and want to confirm the 500-char limit (with and without spaces) at the same time, weighing a single tweet in Korean against X's weighted-280 ceiling, or checking a Smart Store product name (50 chars) against a Naver blog title (40 EUC-KR bytes) against a YouTube title (100 chars) — all in one screen.
Ten platforms at once
| Platform | Unit | Limit | Notes | |---|---|---|---| | Korean resume (incl. spaces) | code point | 500 | Saramin, some Incruit, Korean default | | Korean resume (excl. spaces) | code point − ws | 500 | Jobkorea, some public-sector | | X (Twitter) | weighted | 280 | CJK = 2, URL = 23 | | Instagram caption | code point | 2,200 | Meta official | | Naver blog title | EUC-KR byte | 40 | Safe SERP margin | | Smart Store product name | code point | 50 | Naver rule | | YouTube title | code point | 100 | | | LinkedIn headline | code point | 220 | | | Meta description | code point | 160 | Google SERP truncation | | OG description | code point | 60 | KakaoTalk · Slack · iMessage cards |
Counting units
- Code point — Unicode code point. Hangeul syllables, Han characters, and emoji each count as one.
- Byte (EUC-KR approx.) — ASCII = 1, everything else = 2. True EUC-KR can't encode every character, but this approximation is the standard for Korean SEO usage.
- Weighted (X) — Latin / general punctuation (cp < 0x250 or
0x2000–0x206F) = 1, everything else = 2. URLs (
https?://) count as 23 regardless of length. A simplified twitter-text.
Gauge colors
- Grey — under 80%
- Brand accent — 80–100%
- Red — over the limit
Privacy
Resume drafts, unpublished tweets, private product names, blog outlines —
none of it leaves the page. Counting uses only JavaScript built-ins
([...text].length, TextEncoder, regex); no external libraries.
Not the right tool when
- Body-length statistics (words, sentences, reading time) — that's N°XX text-stats. This tool is for limit checks across SNS, hiring, and SEO platforms.
- Skipping markdown / HTML tags — this counts raw text. Render to plain in markdown preview first, then paste back here.
- Excluding regions (code blocks, comments) — not supported. Everything is counted.
Text stays on the page
Resume drafts, unpublished tweets, private product names, blog outlines — none of it is uploaded. Counting uses just [...text].length, TextEncoder, and regex; zero external libraries.