Title and Description Length in Non-Latin Scripts

Short answer: Search engines cut titles and descriptions by the width they take on the screen, not by a fixed number of characters. A rule such as 60 characters for a title was built from Latin text. Chinese, Japanese and Korean characters are wide, so far fewer fit; Cyrillic and Greek letters are close to Latin; Arabic, Hebrew and Indic scripts vary. Write the meaning first, keep the key words near the start, and test real results in each language.

Many multilingual sites copy one length rule into every language. They check that a title has no more than 60 characters, and a description no more than 160, then publish a translation that was never checked in the real results. In some languages this wastes space; in others it produces titles that are cut off in the middle of the first idea. This guide explains why length is a matter of width, how different scripts behave and a practical way to set limits per language. The English-language basics are in the guide to writing title tags.

Why characters are the wrong unit

Google has said that there is no fixed limit on title length in characters. Titles are shown in a space of limited pixel width, and text that does not fit is cut with an ellipsis. A narrow letter such as i takes far less room than a wide one such as m or w, so two titles with the same character count can occupy very different widths. A tool that counts characters is a rough guide, and it is only roughly right for Latin text.

The width available also changes by device and layout. A title may be shown differently on a phone and on a laptop, and search engines may rewrite it altogether, as covered in the guide to why Google rewrites title tags. For this reason, treat any limit as a planning aid, and put the important words first.

Chinese, Japanese and Korean

Han characters, Japanese kana and Korean hangul are generally displayed as full-width glyphs, roughly the width of two Latin letters or more. As a rough planning figure, a title that fits about 60 Latin characters may carry around 30 CJK characters before it is cut, and a description may carry roughly half the number of characters that a Latin one does. These figures are approximations, and you should check them.

The compensation is that CJK text is dense. A short title can carry a complete idea, since a single character may be a word. A direct translation of an English title is often too long or too sparse. Better to write natively: put the main keyword first, add the brand at the end if there is space and avoid filler. Mixed-script titles that contain Latin brand names or numbers will be narrower in those parts, so measure the real title. For the country-level details, see the Chinese SEO guide, the Japanese SEO guide and the Korean SEO guide.

Cyrillic and Greek

Cyrillic and Greek letters have widths close to Latin ones, although some letters are wider and the average word is often longer than the English equivalent. Translating a title from English into Russian, Ukrainian, Bulgarian or Greek usually gives a text with more characters, so a limit near 60 characters may still be a useful guide, but you may need to shorten the wording. Compound endings and long words can push the key term to the end, where it can be cut off. Rephrase so the main topic appears early.

Arabic and Hebrew

Arabic and Hebrew are written right to left, and Arabic letters connect and change shape depending on position. Average character widths differ from Latin, and the number of characters that fit varies by font. In practice many Arabic titles fit a similar count to Latin, but the variance is large. Check them in the results and in a browser preview. Pay attention to mixed direction, such as an Arabic title that contains a Latin brand name or a number. The direction can look odd in the title if the markup is wrong, and the brand may appear on the wrong side. Use proper direction settings on the page and test the title in a search result, as noted in the guide to right-to-left language SEO.

Hindi, Thai and other scripts

Indic scripts such as Devanagari, Bengali and Tamil, and Southeast Asian scripts such as Thai, use combining marks and have no or few spaces between words in some cases. Character counts can overstate or understate width, since a visible syllable may be built from several code points. Thai is written without spaces between words, so a title that is too long can be cut in the middle of a word, which looks poor. Keep titles short and meaningful at the start, and review real results on mobile devices, where many of your readers will see them. Local guides such as the guide to SEO in India and the Southeast Asia guide discuss script-specific points.

Set a working limit for each language

Instead of a global rule, build a short table that records a planning limit and a hard check for every language. Here is a sensible way to define it:

Check the planning approach against real results: search for your own page with a site query, or look at similar pages from competitors in that language, and see where titles are cut. Update the table when you learn more. The workflow for translated metadata is in the guide to translating title tags and descriptions.

Descriptions and snippets

Descriptions face the same width limit, and they are even less predictable, because Google often builds the snippet from page text instead of using your description, as explained in the guide to how Google picks snippet text. In CJK languages the snippet may show fewer characters but more meaning. Write the description as a clear summary of the page, with the main point in the first sentence. If it is cut, the reader should still understand what the page offers.

Avoid emojis and symbols as a way to gain space, since their width varies and they may be removed. The issue is covered in the guide to emojis and symbols in titles.

Testing in practice

Do not rely on a counter alone. Use a preview tool that measures pixel width with a font similar to the one used in results, and then verify with live results where possible. For each language, test a sample of five to ten pages, including long and short ones, and record how they appear on both phone and desktop. Look especially at pages with brand names, numbers and mixed scripts. If a title is regularly cut, shorten the structure: remove the brand suffix, use a shorter separator or drop a secondary phrase.

Keep a note in your style guide for each language, with an example of a good title and an example that fails. Translators and editors then have a model to follow. Review the notes yearly, since the layout of results changes.

Check also that the title is not too short. A title that uses only a few words wastes the chance to describe the page and invites the search engine to write its own. In dense scripts a very short title can be unclear. Aim for a meaningful phrase that names the topic and, where relevant, the benefit or the type of page. Make titles unique within each language version, because duplicate titles across pages make results confusing, a problem described in the guide to duplicate title tags.

Here is a short example. A software company used the same 60-character rule in all languages. Its Japanese titles were direct translations of about 45 characters, and the results showed them cut after the first third, which removed the product name placed at the end. After a review, the editors wrote native titles of about 25 characters, put the main term first and added the brand only on the home page. Clicks on those pages rose in the following months. The change cost little and came from respecting the width of the script.

A second example is a news site in Arabic and English. The English titles were fine, but Arabic titles with Latin brand names showed the brand on the wrong side in some results. The team corrected the direction markup in the page and shortened a few titles, and the results looked correct after the next crawl. The lesson is that a length rule is only one part of the checks; direction and mixing matter too.

Site SEO AI Audit crawls your site and reports titles, descriptions, duplicates, language attributes and hreflang for each page, so you can find missing or repeated metadata in every language. It does not measure pixel width or preview how titles look in results. The first audit of up to 200 pages is free; larger sites can see the pricing page.

Related reading

• Meta Description Writing Guide
• Transliteration and SEO
• Multilingual Keyword Research

The bottom line

Titles and descriptions are cut by pixel width, so set a planning limit per script, put the main idea first and verify each language in real results.

GYIK

Is 60 characters a rule for every language?

No. It is a Latin-text guideline. Pixel width decides, and wide scripts such as Chinese fit far fewer characters.

How long should a Japanese title be?

As a rough plan, about half the character count of a Latin title, then check real results.

Do I need different descriptions in each language?

Yes. Write each natively and put the main point first, because snippets are often cut or rewritten.

Why does my Arabic title look odd?

Mixed direction with Latin brand names can misplace words. Check direction markup and test in results.

Can I use symbols to save space?

It is unreliable, since widths vary and symbols may be removed from displayed titles.