Count symbols online
Paste your text or upload a file. Four different totals, a breakdown of every character, the hidden ones nobody typed, and how many SMS segments it costs.
Paste your text and it counts the symbols four ways, because four different systems count differently and all four are correct. It also breaks the text down by kind of character, lists every distinct one with its Unicode code point, finds the invisible characters nobody typed, and works out how many SMS segments the message costs. Free, no sign-up, nothing leaves your browser.
Three steps, no sign-up, no upload.
Type, paste, or drop a .docx, .pdf or .txt file onto the editor. Files are read in your browser; nothing is uploaded.
Symbols, characters, code points or bytes. The table below says which one a given limit means.
The Hidden tab finds non-breaking spaces, zero-width characters and soft hyphens, and removes them in one click.
Every one of these numbers is correct. They are answers to different questions, and the reason a character count sometimes looks wrong is that you were given the wrong one. Every row below was measured with this tool before it was published.
| You type | Symbols | Characters | Code points | UTF-8 bytes |
|---|---|---|---|---|
Hello — plain Latin | 5 | 5 | 5 | 5 |
café — é as one character | 4 | 4 | 4 | 5 |
café — é as e + accent mark | 4 | 5 | 5 | 6 |
| 👍 thumbs up | 1 | 2 | 1 | 4 |
| 👨👩👧 family | 1 | 8 | 5 | 18 |
| 🇵🇰 flag | 1 | 4 | 2 | 8 |
नमस्ते — Hindi | 3 | 6 | 6 | 18 |
The two café rows are the same word and look identical on screen. One was typed with an accented é; the other with a plain e followed by a combining accent. A form that allows 4 characters accepts one and rejects the other, and nothing on screen tells the person which they pasted.
A real run, recorded on 19 September 2026, on the sample message behind the “Try an example” button — a 133-character delivery notification of the kind businesses send every day.
133 symbols, 134 characters, 133 code points, 139 bytes. 110 characters once spaces are excluded. By kind: 90 letters, 24 spaces, 10 digits, 8 punctuation marks and 1 emoji — plus one non-breaking space that is invisible, was pasted in, and is counted by everything.
Text pasted from Word, a PDF or a web page brings passengers. They are invisible, they count against every limit, and they break exact-match search and comparison tools without leaving a trace.
Upgrade
Drop your file inU+00A0Looks exactly like a space and never wraps. Arrives from Word and from HTML . The most common hidden character by a wide margin, and the usual reason a search for a phrase finds nothing.U+200BOccupies no space at all. Used legitimately to hint where a long word may break; otherwise pure noise that silently splits a word in two as far as any search is concerned.U+00ADA hyphen that only appears when a word wraps. It survives copy and paste, so it travels into your text and stays invisible until something else renders it.U+FEFFA file marker that ends up inside the text when a file is saved or converted carelessly. Shows up as a stray character at the start of a CSV or a JSON file.U+200E, U+200FLeft-to-right and right-to-left controls. Legitimate in mixed Arabic, Hebrew, Urdu and Persian text; accidental everywhere else, and counted against your limit either way.SMS does not count characters the way anything else does, which is why a message that looks like 150 characters can arrive as two.
The standard SMS alphabet packs seven bits per character, so a single message holds 160. Once it splits, each part holds 153 — the rest is used to stitch the parts back together.
One character outside the GSM alphabet forces the whole message into UCS-2 and halves the capacity to 70, or 67 per part. It is all or nothing: there is no partial upgrade.
A curly apostrophe, an em dash, an ellipsis character, a non-breaking space, any emoji. Also worth knowing: [ ] { } \ ^ ~ | and the euro sign are in the alphabet but cost two units each.
The SMS tab names the exact characters forcing your message into UCS-2, and how many of each. Most of the time the fix is a straight apostrophe and a hyphen.
Three jobs the four numbers matter to.
Applications, bios, ad copy, product fields. Paste the text, read the character count, and — if the form is rejecting something that looks short enough — check the Hidden tab, because an invisible character is usually the answer.
Notifications, reminders, marketing. One stray em dash across fifty thousand messages is fifty thousand extra segments, and nobody notices until the invoice arrives.
Working out why a string that fits in one system is rejected by another. Code points against code units against bytes, with every distinct character listed by code point, usually settles it in seconds.
What people ask when they search for a way to count symbols online.
Paste your text into the box above. The symbol count appears immediately, next to three other totals that are also correct and also different: characters, Unicode code points and UTF-8 bytes. Which one you need depends on what is doing the limiting, and the table above says which is which.
In most languages that ask this question — Russian символ, Polish znak — "symbol" simply means character. So a symbol counter is a character counter. The wrinkle is that one thing a reader calls a symbol can be several characters underneath, which is why this page shows four numbers instead of one.
Because four different systems count differently and all of them are right. A thumbs-up emoji is 1 symbol, 2 characters in a browser form field, 1 Unicode code point and 4 bytes in UTF-8. A family emoji is 1 symbol, 8 characters, 5 code points and 18 bytes. Any tool showing one number is picking one of those and not telling you.
For a text box or a form limit, the character count — that is what JavaScript measures. For a database column or an API field, code points. For SMS, the SMS tab, because SMS counts in its own units. For anything a person is going to look at and count by eye, symbols.
Both. The total and the figure excluding all whitespace sit together on the Counts tab and in the footer, because forms disagree about which they mean and you should not have to guess.
Non-breaking spaces, zero-width spaces, soft hyphens, byte order marks and direction marks: characters you cannot see but that count against every limit, break exact-match search and show up as mystery characters in other software. They arrive with text pasted from Word, PDFs and web pages. The Hidden tab finds them and one button removes them.
Only the invisible ones. Each is replaced with what it should have been — a normal space, a line break, or nothing at all — and any double spaces left behind are collapsed. Your visible text is untouched, and it never leaves the box.
160 when every character is in the GSM 03.38 alphabet, or 70 when it is not. Once a message splits, each part carries 153 or 67, because the rest is used to stitch the parts back together. The SMS tab shows which alphabet your message is in, why, and how much room is left.
Because it is outside the GSM alphabet, which forces the entire message into UCS-2 and halves the capacity. A curly apostrophe, an em dash or an emoji will do it. On the example above, three such characters turn a 130-character message into two billed segments; removing them brings it back to one.
It counts them four ways and shows you all four, which is the only correct answer. Modern browsers segment emoji properly, so a flag or a family counts as one symbol. Older browsers without grapheme segmentation fall back to code points, and the page says so when that happens.
Yes. The Breakdown tab lists every character that appears, with its Unicode code point, its kind and how many times it occurs, and you can filter by character, by kind or by a code such as U+00A0. It is the quickest way to find the one strange character in a long file.
No. Every count runs in your browser, with no server call carrying your text. That is also why the tool is free, unmetered and works on a pasted password reset email or a draft contract without you having to think about it.
The word counter is about words, reading time, readability and platform limits. This is about the characters themselves: how many there are under four different definitions, what they are made of, which of them are invisible, and what they cost to send. If you want a character count to check against a limit, either page will do; if a character count looks wrong, this is the page that explains why.
Counting the characters is one job. These are the others.
Free and unmetered, because it runs in your browser. Plan allowances apply only to the AI tools, and they are on the pricing page.