40.9%
English share of web pages, 2026
▼ 2.8 pts since 2019
6.0%
German share of web pages, 2026
▲ 0.5 pts since 2019
4.7%
Chinese share of web pages, 2026
▼ 0.2 pts since 2019
0.6%
Arabic share of web pages, 2026
▼ 0.1 pts since 2019

Why pages, not homepages.

Most reports on "languages of the internet" come from W3Techs, which detects the language of the homepage of each of the top 10 million websites. That's a useful view but it counts a multilingual site like apple.com or microsoft.com as a single English site, even when much of their content is multilingual.

This dashboard uses Common Crawl instead: the language of every single page across roughly three billion pages per crawl. It's the same data the major large language models (GPT, Claude, Gemini) are pre-trained on. For localization strategy and AI search optimization, pages are what matter. That's what AI reads, indexes and answers from.

Side note: The two methodologies don't fully agree. W3Techs reports English at ~49 % and a sharp decline since 2020; Common Crawl puts English at ~43 % and roughly stable. The difference is what gets counted as "the web." I've chosen the page-level view because it's the one your audience actually reads and your AI tools actually train on.

Eight years on the line.

Annual share of pages crawled by Common Crawl, by content language, 2019 to 2026. Toggle a language to add or remove it from the chart.

Y-axis on a logarithmic scale so the 0.06–46 % range reads on one chart. Source: Common Crawl monthly archive statistics (CC-MAIN-YYYY-WW). Earliest crawl of each year. Each crawl covers ~3 billion pages; language is detected per page via Compact Language Detector 2 (CLD2).

Where every language stands.

The 29 most-published content languages on the open web, ranked by their 2026 share. The original 25 plus four large-speaker languages — Hindi, Bengali, Thai and Malay — added in June 2026 to surface the most under-read languages in AI training data. Click a column to sort; type to filter.

# Language 2019 2020 2021 2022 2023 2024 2025 2026 Δ 2019→26 Trend

Sparklines show each language's 2019–2026 trajectory at a common vertical scale (0%–50%). Source: Common Crawl.

The web is not the world.

For every language, I compare its share of web pages (Common Crawl, 2026) with its share of the world's speakers according to Ethnologue, L1 (first language) + L2 (second language). Bars to the right of zero indicate languages overrepresented online; bars to the left, underrepresented.

Note: If we consider only speakers for whom English is a first language, the figure drops from 1,493M to 372M (Ethnologue, 2026). In that case, English’s overrepresentation would be even more pronounced.

Speaker share = (L1 + L2 speakers, Ethnologue 29th edition) ÷ 8.2 billion world population. Chinese, Hindi, Arabic and Bengali are highlighted; all four are vastly underrepresented online relative to the number of people who speak them.

Web share vs. speaker share.

A direct comparison. The gap column is web % minus speaker %. Negative values mark languages whose web presence trails their speaker base.

Language Speakers (M, L1+L2) Speaker share of world Web share, 2026 Gap (web − speaker) Status

How well AI reads each language.

This is a single ratio: a language's share of web pages divided by its share of the world's speakers. Below 1.0, AI under-reads the language relative to its speaker base; the localization and AEO opportunity is largest. Above 1.0, the language is over-represented in the training data; quality is high and the marginal payoff to adding more content is lower.

Read the bars on a log scale. Hindi and Bengali sit at ≈0.03 (AI sees them at 3 % of their speaker share). Arabic at 0.16 is severely under-read; Czech at 8.9 means AI sees Czech at almost nine times its speaker share. The dashed line at 1.0 marks parity. Languages with an index of 4 or higher reflect very prolific publishing cultures relative to their population: the European bloc, plus Japanese.

In summary

Seven things to take from the data.

  1. English's late-cycle slip is real. Page-level English was flat between 44 % and 46 % for five straight years (2019–2023), then dropped to 43.4 % in 2025 and 40.9 % in 2026, five points shed in two crawls. The cause is non-English content rising faster than English, not a sudden English decline.
  2. The web has several near-peers. German, Japanese, French and Spanish all sit between 4.7 % and 6.4 % of pages in 2026. Closer to each other than any of them is to English above or to the long tail below.
  3. Hindi and Bengali are the most under-read major languages on the web. Hindi's AEO index is 0.03 (610M speakers, 0.22 % of pages); Bengali is also 0.03 (278M speakers, 0.11 % of pages). AI sees these two at roughly 3 % of the rate they're actually spoken — five to six times more extreme than Arabic, which used to be the worst at 0.16. Add Malay (0.08) and Tamil/Urdu (off the chart but below 0.05) and the picture is clear: Indic and Austronesian languages are where the localization and AEO opportunity is largest.
  4. Europe over-publishes, by a lot. Czech, Danish, Dutch, Swedish, Polish, German plus Japanese all show AEO indices above 3.5. AI reads them at three-to-nine times their speaker share. A productive open-web publishing culture, not population, drives saturation.
  5. The Central and Eastern European bloc is rising in concert. Polish (+0.3 pts), Ukrainian (+0.4 pts), Turkish (+0.4 pts) and Czech (+0.1 pts) have all gained meaningful page share since 2019. Four neighbouring language markets growing together as content production accelerates.
  6. Ukrainian has more than doubled since 2019. 0.38 % → 0.81 % of pages, the steepest rise in the top 25. Massive content production after 2022 has put Ukrainian materially above its speaker share for the first time.
  7. Composition is sticky, ranks are not. Inside the top 25 measured continuously since 2019, no language has dropped out. What changed is the ordering. The four languages added in June 2026 (Hindi, Bengali, Thai, Malay) were always below the 0.25 % web-share threshold, but they're the ones where the speaker-to-web gap is largest — the stubborn floor where the biggest under-represented populations live.