Character Counter

Where character limits actually bite

Counting characters only matters because something downstream truncates, rejects, or splits your text. Those thresholds are scattered across platforms and rarely documented where you need them, so here they are in one place — along with the reasons a count can disagree with the one a platform shows you.

Common limits worth knowing

• SMS — 160 characters in the basic GSM alphabet. Include one character outside it (an emoji, a curly quote, most accented letters) and the whole message switches to Unicode encoding and the limit drops to 70. Longer messages are split and sent as segments of 153 or 67 characters, and you are billed per segment.
• Meta description — no hard limit, but Google truncates the displayed snippet at roughly 155–160 characters on desktop and less on mobile. The cut-off is measured in pixels rather than characters, so wide letters shorten it further.
• Title tag — around 60 characters before truncation, again pixel-based.
• Email subject lines — no technical limit, but mobile clients commonly show only 35–45 characters. Front-load the meaning.
• Search ad headlines — 30 characters each, hard-enforced, which is the tightest constraint most writers meet.
• Alt text — no formal limit, but some screen readers cut off around 125 characters, so treat that as a working ceiling.

Social platforms change their limits often enough that any number printed here would eventually mislead. Check the platform's own composer, and use a counter to draft against whatever it currently enforces.

Why your count and the platform's count disagree

"Character" is not one thing, and the difference shows up the moment emoji or accents are involved.

• Code units. Many systems count UTF-16 units, where most emoji occupy two. An emoji that looks like one character often counts as two.
• Bytes. Databases and APIs frequently limit bytes, not characters. In UTF-8 an ASCII letter is 1 byte, an accented Latin letter is 2, most CJK characters are 3, and emoji are 4. A 255-byte field might hold 255 English characters or only 63 emoji.
• Grapheme clusters. What a reader sees as one character can be several code points combined — a family emoji built from several people joined by zero-width joiners can count as 7 or more, and an accented letter may be a base letter plus a combining mark rather than a single precomposed character.

The practical consequence: if you are near a limit and your text contains emoji, accented characters, or non-Latin scripts, paste it into the actual composer before publishing. A counter tells you the shape of the problem; only the target system knows its own rule.

Words, characters, and reading time

Word counts suit essays and articles; character counts suit anything with a hard container. Reading time estimates sit on top of the word count, and the figures used almost everywhere are 200–250 words per minute for silent reading and about 150 for reading aloud — so a 1,000-word article is roughly four to five minutes silently and closer to seven spoken.

Those averages hide a lot. Dense technical text is read far more slowly than narrative, and readers skim rather than read on screen. Treat a reading-time figure as a rough signal of length, not a measurement.

Reading keyword density honestly

Density is simply how often a term appears as a share of total words. It is useful as a check against accidental repetition, and it is not a target to optimize toward. No modern search engine rewards hitting a particular percentage, and writing to one produces text that reads badly, which does hurt.

The useful way to read a density panel is as a summary of what your page is about. If the top terms are not the subject you meant to write about, that is worth knowing. If your main topic does not appear at all, the page may be vaguer than you think. Beyond that, write for the reader.

Whitespace, line breaks and other invisible characters

Counts differ depending on what is included, which is why two tools can disagree on identical text:

• Trailing spaces and blank lines at the end of a paste are counted by most tools and are easy to miss.
• A line break may be one character or two, depending on whether the text came from Windows, and pasting into a form can convert between them.
• Text copied from a web page or PDF often contains non-breaking spaces, which look identical to normal spaces but behave differently in searches and comparisons.
• Word processors substitute curly quotes, em dashes and ellipses. A single ellipsis character is one character here and three in a system that expands it, and curly quotes are enough to push an SMS into Unicode encoding.

Writing to a tight limit

• Write the full version first, then cut. Composing inside the limit produces cramped text.
• Cut qualifiers before content — "really", "very", "in order to", "it is worth noting that". These are the cheapest characters on the page.
• Put the essential words first. Whatever truncates will take from the end.
• Shorten links properly rather than trimming words, if links count toward your limit.
• Check the result in the destination, not just the counter. Line wrapping and pixel-based truncation are not visible in a plain count.

Privacy

Everything you type or paste here is counted in your browser. The text is not uploaded, not logged, and not visible to anyone running this site — which matters if you are drafting something confidential.

Frequently Asked Questions