Character counter with and without spaces
Count characters with or without spaces and compare two drafts using the same rule. This counter separates visible characters, Unicode code points and UTF-16 units to explain differences caused by emoji and combining accents. Everything runs on your device, with instant results, free exports and no account or email required.
Your texts and results
- Visible characters (graphemes)
- 0
- Unicode code points
- 0
- UTF-16 units
- 0
- Characters without whitespace
- 0
- Words
- 0
- Approximate sentences
- 0
- Paragraphs
- 0
- Lines
- 0
- Letters (code points)
- 0
- Decimal digits
- 0
- UTF-8 bytes
- 0
- Unique words
- 0
- Code points per word
- 0
- Visible characters (graphemes)
- 0
- Unicode code points
- 0
- UTF-16 units
- 0
- Characters without whitespace
- 0
- Words
- 0
- Approximate sentences
- 0
- Paragraphs
- 0
- Lines
- 0
- Letters (code points)
- 0
- Decimal digits
- 0
- UTF-8 bytes
- 0
- Unique words
- 0
- Code points per word
- 0
One shared constraint
Import a local document
Choose a TXT, MD, DOCX or ODT file up to 2 MB. Only the main text is extracted; formatting, headers, footnotes, images and tracked changes are not reconstructed. Check the extracted text before counting. Your document is never uploaded. Encrypted files and unknown formats are rejected.
How to use the counter
Paste both drafts or import a document. Select the rule your recipient requires and enter your own limit. Counts and the remaining allowance update while you edit. A negative allowance means you exceeded the limit; B minus A shows whether the second draft is longer. Undo and redo cover the last twenty states of each field during this visit. Export statistics to keep a comparison, and copy the text separately if you want to share it.
Method: what counts as a character?
Text is preserved except CRLF and CR line endings become LF. Graphemes use Intl.Segmenter and represent groups of code points perceived as one visible character. Code points are counted by Unicode iteration; UTF-16 units use the string’s technical length. The whitespace-free total removes spaces, tabs and line breaks recognized by JavaScript before segmentation. Letters and digits are code points in Unicode categories L and Nd. Words use your browser’s language segmentation; unique words ignore case only. Blank lines separate paragraphs and LF separates lines. Sentence counts approximate boundaries using . ! ? and their CJK equivalents, without grammatical analysis. Reading and speaking estimates divide words by your chosen personal pace; they do not promise any speed.
Examples and limits
Example: Bonjour le monde ! contains 18 graphemes, 18 code points and 18 UTF-16 units, including 15 non-whitespace characters and 3 words. The family emoji 👨👩👧👦 is usually 1 grapheme, but 7 code points and 11 UTF-16 units. The letter é may be precomposed (1 code point) or written as e plus a combining accent (2 code points), while remaining 1 grapheme. Literal search can ignore case or accents; matches do not overlap. A sequence of characters is searched as a sequence, never as a regular expression.
Your recipient’s platform may use another rule, omit some whitespace or count links differently. Always check its final preview. Limits here are your settings, not current social platform quotas. Each field accepts up to 100,000 UTF-16 units. When Intl.Segmenter is unavailable, the tool replaces graphemes with code points and displays a notice; do not then rely on the visible-character count for complex emoji. Segmentation may vary with the browser and Unicode version. This tool does not score quality, originality or the personal value of your writing.
Privacy
This tool sends no entered text or document. Drafts remain in tab memory and are never saved by our buttons. Save settings stores only the limit, counting rule and reading/speaking rates in this browser; Delete settings removes them. CSV and text exports include statistics, never drafts or the search phrase. CSV headers are localized and numbers use a decimal point. Copying a draft is a separate, deliberate action. Reload or close the tab to clear its history; remember the clipboard and downloaded files when using a shared device.
Frequently asked questions
Do spaces and line breaks count?
Yes in the full totals. The whitespace-free total removes them, including tabs and nonbreaking spaces recognized by JavaScript.
Why can an emoji count more than once?
One visible sign may combine several code points and UTF-16 units. The panel shows all three counts without assuming platforms use the same rule.
Does the tool change my accents?
No in the main counters. Only line endings are normalized. The accent option applies exclusively to occurrence searches.
Can I count a Word document?
Yes, unencrypted DOCX files are read locally. Only the main text is extracted. Check tables, footnotes and line breaks in the field after importing.
Can I retrieve a draft after closing?
The tool does not save drafts. Copy your text into your own document before closing. Saved settings contain no draft.
Will this equal an SMS or social platform limit?
Not necessarily: encoding, links, emoji and service rules can change the count. Select the requested convention and check the destination service.
Sources and conventions
Consulted November 14, 2026: Unicode UAX #29 for grapheme and word boundaries; ECMA-402 for Intl.Segmenter; WHATWG Encoding for TextEncoder. These standards do not supply writing scores or editorial thresholds.