Twitter's 140-Character Limit Was Set by SMS Arithmetic in 2006. Here Is What Character Limits Actually Measure.

ToolHQ TeamAugust 4, 20267 min read

In November 2006, Jack Dorsey posted the first tweet: "just setting up my twttr." The character limit was 140, chosen because SMS messages are limited to 160 characters and Dorsey's team wanted to leave room for a username. A 160-character SMS limit became 140 characters of content plus 20 characters for a handle and colon prefix. That was the entire reasoning.

For a decade, 140 characters was the most consequential character limit in digital communication. Users learned to write within it. Abbreviations became conventions. The URL shortener industry exists largely because of it. Twitter doubled the limit to 280 in November 2017, citing data showing that roughly 9 percent of English-language tweets hit the 140-character ceiling while less than 1 percent of Japanese tweets did, because ideographic characters convey more information per symbol. The original constraint shaped how millions of people learned to write online before the extension.

Understanding where character limits come from, what they actually measure, and why the count depends on which counting method you use makes character counter tools more useful and character limit decisions more intentional.

The History of Character Limits in Communication

The concept of fixed character limits in written communication predates digital media. Telegraph operators charged by the word, which created economic pressure toward brevity and produced the compressed telegraphic style: "SHIPMENT DELAYED STOP ARRIVAL THURSDAY STOP." This compression influenced early twentieth-century journalism, where telegraphed dispatches to newspaper wire services were billed by word count and editors enforced strict length requirements.

SMS messages are limited by the Global System for Mobile Communications standard, finalized in 1985. The GSM specification allocated 140 bytes for the message payload in each SMS unit. Using the 7-bit GSM character encoding, which covers the Latin alphabet and common punctuation, each character requires 7 bits, allowing 140 times 8 divided by 7, which equals 160 characters per message. When messages include characters outside the GSM-7 set, the encoding switches to UCS-2, a 16-bit encoding, which reduces the per-message limit to 70 characters.

Twitter's designers, building in 2006 when SMS was the primary mobile messaging medium, designed around the SMS constraint. The intent was that a tweet could be sent and received as a text message, with the sender's username prepended automatically. A typical Twitter username of around 15 to 17 characters plus the "@" sign and a colon consumed approximately 20 characters of the 160-character budget, leaving 140 for the message content. Twitter's connection to SMS as a delivery channel made this constraint feel natural even as smartphones with internet connectivity made SMS-based tweet delivery largely unnecessary.

What Character Count Actually Measures

A character count sounds simple until you encounter text that contains more than the Latin alphabet. The ambiguity begins with what we mean by "character."

In ASCII, the encoding standard developed in 1963 for English-language text, each character is one byte and one visual symbol. The count of bytes, code units, and visual characters all agree. ASCII covers 128 characters including the English alphabet, digits, punctuation, and control characters. This simple case is what most character counter tools were originally designed for.

Unicode, the modern standard for representing text in all human writing systems, assigns code points, numerical values, to more than 149,000 characters across 161 scripts as of Unicode 15.1. Code points are encoded into bytes using encoding schemes: UTF-8, which uses one to four bytes per code point; UTF-16, which uses two or four bytes; and UTF-32, which uses exactly four bytes. In UTF-8, ASCII characters are still one byte, but characters from other scripts require more.

The complication deepens with combining characters and grapheme clusters. A letter with a diacritic, such as the letter e with an acute accent, can be represented in Unicode as either a single precomposed code point (U+00E9, one code point) or as a base letter plus a combining accent mark (U+0065 followed by U+0301, two code points). Both produce an identical visual character. A character count that counts code points gives different results depending on which representation is used.

Emoji add another layer of complexity. Many emoji are single code points: the smiley face is U+1F600. But many others are composed of multiple code points joined by zero-width joiners. The family emoji showing a man, woman, and child is actually three individual emoji characters (U+1F468, U+1F469, U+1F466) connected by two zero-width joiner characters (U+200D). That single visible symbol consists of five Unicode code points. A character counter that counts code points reports five; one that counts visual grapheme clusters reports one; Twitter's counting system historically counted emoji as two characters regardless of complexity.

Where Character Counts Have Real Consequences

Search engine metadata is one context where character counts carry direct business impact. Google displays page titles up to approximately 50 to 60 characters and meta descriptions up to approximately 150 to 160 characters before truncating with an ellipsis. The actual cutoff is determined by pixel width, not character count. Google's search results page allocates a fixed pixel width for snippets. Wide characters like capital W, M, and D consume more pixels per character than narrow ones like i, l, and j.

This means that a 60-character title composed primarily of wide capital letters may be truncated while a 65-character title with mostly narrow lowercase letters displays completely. Professional SEO writers who optimize titles to exactly 60 characters are working with an approximation that assumes average character width. The practically usable rule is to keep titles under 55 to 60 characters with a visible preference for titles that end at a natural phrase boundary before 58 characters, where truncation almost never occurs.

Advertising platforms have precisely defined character limits that are technically enforced rather than approximately followed. Google Search Ads allow headlines of up to 30 characters and description lines of up to 90 characters. Responsive ads that exceed these limits cannot be saved. Facebook advertising headlines are limited to 40 characters, primary text to 125 characters, and link descriptions to 30 characters. These are hard constraints rather than guidelines, and copy that exceeds them must be revised before the ad can run.

LinkedIn's advertising character limits differ again: 70 characters for headline in Sponsored Content, 150 characters for description. Twitter's promoted tweet format follows the same 280-character limit as organic tweets. Each platform's limit reflects its interface constraints and content format, not a shared standard.

The Practical Contexts for Character Counting

Email subject lines have no technical character limit in the SMTP protocol that governs email delivery, but email client interfaces impose visual limits. Gmail's inbox list view displays approximately 60 to 70 characters of a subject line on a desktop. Mobile email apps display fewer, often 30 to 40 characters before truncation. Research on email open rates consistently correlates subject lines of 40 to 60 characters with higher open rates, likely because they complete a thought within the visible preview without being truncated.

Push notifications from mobile apps have stricter limits: iOS displays approximately 100 to 120 characters before truncating in the notification center, with shorter limits in banner notifications while the phone is in use. Android's notification length varies by manufacturer's notification tray implementation but typically shows 40 to 80 characters in the banner and more in the expanded view. Notification copy written to deliver its key message in the first 50 characters performs consistently better than copy that front-loads context.

Conclusion

Code comments and documentation have no formal character limits but are typically subject to line length conventions. The Python style guide PEP 8 recommends lines of 79 characters or fewer. Many code style guides enforce 80 or 100 character line length limits. These limits derive from the early days of terminal displays that showed 80 characters per line, a convention from IBM's 80-column punch card format that predates video terminals.

Understanding which counting method a given platform uses, whether it counts bytes, Unicode code points, UTF-16 code units, or visual grapheme clusters, and knowing the effective limit in those terms before writing copy, is what makes character counting a precision tool rather than an approximation.

Frequently Asked Questions

Why did Twitter originally have a 140-character limit?

SMS messages are limited to 160 characters. Twitter's founders wanted space for a username prefix, which they estimated at 20 characters, leaving 140 for content. The limit was a practical arithmetic decision, not a philosophy about brevity.

Does an emoji count as one character?

It depends on the system. Some platforms count each Unicode code point, and many emoji use multiple code points. Others count visual grapheme clusters, where a complex emoji counts as one. Twitter counts emoji as two characters.

How many characters should a meta title be?

Google recommends keeping meta titles under 60 characters as a guideline, but the actual cutoff is based on pixel width, not character count. Wide letters like W take more pixels than narrow ones like i.