Character counter
One text, every measure that matters: perceived characters (graphemes), the UTF-16 units some platforms count, UTF-8 bytes and words — each against the real limit of the platform you're writing for, with its source and verification date stated.
“Characters” counts perceived letters (graphemes): one emoji, even one made of several code points, is one character. Platforms count different measures — the UTF-16 and byte rows tell you what a strict counter would say.
What one emoji costs: X — 2 for almost every emoji, 4 for flags, and ZWJ sequences charge each part, so 👨👩👧👦 = 11 (official weighted ranges). Threads — an emoji costs its UTF-8 bytes (official): 😀 = 4, the family above = 25. TikTok — full-width characters count 2 only in the documented ad rules; captions have no official figure. Bluesky, Discord, YouTube, LinkedIn and Instagram work by graphemes: every emoji is 1, exactly the measure counted here.
Verified 2026-10-23 against each platform's own documentation where one exists (sources and capture hashes in the table below). “Reported” limits are widely confirmed in practice but not published by the platform; TikTok's caption has no official figure — its 2,200 bar is the conservative working figure (the API/external-tools number; the app can reach 4,000).
Every limit, with its source
Checked 2026-10-23. Each official limit below was captured from the platform's own documentation — the quotes, capture dates and file hashes are archived with this project's reference material. Character limits change without notice; every row names the exact source to re-check. The figures themselves are the bars above — this table is deliberately slim provenance: where, how official, and the exact source.
| Where | Status | Source |
|---|---|---|
| X post | Official | docs.x.com — Counting characters |
| Threads post | Official | Meta — Threads API, Posts |
| Bluesky post | Official | Lexicon app.bsky.feed.post (maxLength / maxGraphemes) |
| YouTube title | Official | support.google.com/youtube |
| YouTube description | Official | support.google.com/youtube |
| Discord message | Official | discord.com/developers — Message object |
| LinkedIn post | Official | LinkedIn Help — Post and share updates |
| Instagram bio | Official | help.instagram.com — Add a bio |
| Instagram caption | Reported | Widely confirmed in practice; Meta's own docs omit it |
| TikTok bio | Reported | Widely confirmed in practice; TikTok's Help Center omits a published figure |
| TikTok comment | Reported | Widely confirmed in practice; no official page publishes it |
| TikTok caption | Unconfirmed | Third-party accounts: 2,200 via the API or external tools, up to 4,000 in the mobile app (varies by version and post type) — TikTok publishes nothing official |
| TikTok ad caption | Official — ads only | ads.tiktok.com — In-Feed Ads counting rules (captured 2026-10-23) |
| Meta description | Official — no limit stated | Google Search Central — Snippets |
Graphemes, UTF-16 and bytes are different counts
"Character" is ambiguous until you pick a definition. A grapheme is what a reader perceives as one character: ñ is one, and an emoji built from several code points — even a family emoji joined with zero-width joiners — is one grapheme. UTF-16 code units are what JavaScript's .lengthreports (that emoji is 2); UTF-8 bytes are what networks transmit (that emoji is 4). This tool labels all three instead of pretending one number covers them.
Bluesky is the platform whose official limit is defined in graphemes (300, or 3,000 bytes) — its own documentation separates the two. X is the other special case, and the widget above implements its official rules: the 280 limit is aweighted count where every emoji counts 2 no matter how complex, CJK characters count 2, any URL counts 23 (t.co shortening) and the text is normalized to NFC first — the same weights X's own counting configuration defines. Two X edge cases can't be detected in a browser (attached media and auto-populated reply mentions), so the X row can only overestimate, never undercount. Threads applies its documented emoji rule: each emoji costs its UTF-8 bytes, everything else counts 1.
What is not official — and why that matters
Four rows are deliberately not presented as facts. Instagram's 2,200-character caption limit is confirmed everywhere in practice, but neither Instagram's Help Center nor Meta's developer documentation publishes it today — we checked, and the checked absence is documented. TikTok's organic caption has no official figure either, and third-party accounts agree more than they conflict: 2,200 characters when posting through the API or external tools, up to 4,000 from the mobile app — it varies by app version and post type (photo posts being the classic example); the caption bar follows the conservative half of that — 2,200, the number API posting and every scheduling tool enforce. The same pattern holds for TikTok's bio (~80) and comment (~150) figures: stable in every third-party guide, never published by TikTok itself. The one TikTok number stated as official here applies to ads captions: TikTok's own ads help center documents max 150 characters displayed, full-width CN/JP/KR characters count 2, and leading and trailing spaces are skipped — captured byte-exact on 2026-10-23. When a platform won't state a number, a good tool says so instead of dressing an estimate as a limit.
FAQ
Do hashtags count toward the limit?
On the platforms that document it, yes — the limit covers everything in the text, hashtags included (LinkedIn states it explicitly). Visible truncation is a separate matter: feeds show only the first ~125 characters (Instagram, TikTok) or ~140–210 (LinkedIn) before a "more" link, so front-load what matters.
Why does my counter disagree with the platform?
Usually because it counts a different measure (UTF-16 vs graphemes), because URLs are expanded or shorted on paste, or because the platform applies weighted rules (X). Compare this tool's three counts against the platform's own counter once and you'll know which measure it uses.
Is anything sent to a server?
No. The counting runs entirely in your browser with the standardIntl.Segmenter API.
More text tools
- Text styles — 15 Unicode styles with Spanish-safe substitution
- Character counter — graphemes, UTF-16, words and every platform limit
- Case converter — UPPER, lower, Title, camelCase, snake_case and more
- Text cleaner — spaces, line breaks, duplicate lines, invisible characters
- Emojis — search and copy the official Unicode 17.0 set
- All text tools