HTML was created by Tim Berners-Lee in 1991 as part of his proposal for the World Wide Web, and the first website went live that same year. In the more than three decades since, HTML has evolved from a handful of simple tags into a formally standardized language maintained by WHATWG — and it remains, unchanged in its core purpose, the language browsers use to structure and render every page on the web.

What is HTML?

Before tracing the timeline, a brief anchor: HTML (HyperText Markup Language) is the markup language that describes the structure of a web page — headings, paragraphs, links, images — which a browser then renders visually. For a fuller explanation, see our guide to what HTML is.

1989: The World Wide Web is proposed

In March 1989, Tim Berners-Lee, then a researcher at CERN (the European Organization for Nuclear Research), circulated a proposal titled "Information Management: A Proposal." It described a system for linking and sharing documents across a network using hypertext — text containing links to other text — to solve a real problem: researchers at CERN struggled to keep track of and share information across different systems and formats. This proposal became the conceptual foundation for the World Wide Web.

1991: HTML and the first website

Berners-Lee followed the web proposal with an implementation: HTML as the markup language, HTTP as the protocol for transmitting it, and a browser to render it. In 1991, the first website went live at CERN, and Berners-Lee shared a document describing HTML's early tags on a public mailing list. That document effectively became the first HTML specification — informal by later standards, but the direct ancestor of every HTML page since.

HTML 2.0 and early standardization

As the web grew beyond CERN through the early 1990s, the need for a shared, formal specification became clear — different browsers were beginning to interpret tags inconsistently. The Internet Engineering Task Force (IETF) published HTML 2.0 in 1995, consolidating the tags already in common use into the first official standard, so that pages would behave predictably across different browsers.

HTML 3.2 and HTML 4

Stewardship of HTML moved to the newly formed World Wide Web Consortium (W3C), which released HTML 3.2 in 1997 and HTML 4 in 1997–1999. These versions added tables, richer text formatting, and support for stylesheets — the beginning of separating a page's structure (HTML) from its presentation (CSS), a division that shaped how the web has been built ever since.

The road to HTML5

By the early 2000s, the W3C had shifted focus toward XHTML, a stricter, XML-based version of HTML. In practice, XHTML's strictness didn't match how the web actually worked — browsers and page authors wanted more flexibility, not less. In 2004, a group of browser vendors formed WHATWG (the Web Hypertext Application Technology Working Group) to develop HTML in a different direction: driven by real-world compatibility and practical use cases rather than formal XML purity. That work became HTML5.

HTML5 and modern web development

HTML5, developed through WHATWG and later co-published with the W3C, introduced native support for audio and video, new semantic elements (<header>, <article>, <nav>, and others that describe a page's structure more meaningfully than generic <div>s), and a wide set of APIs that let HTML support far richer, more interactive applications than the original document-oriented language ever anticipated. HTML5 is the version most people mean today when they simply say "HTML."

HTML today

HTML is no longer versioned in discrete numbered releases. WHATWG maintains it as a Living Standard — a single, continuously updated specification that evolves as browsers and the web's needs change, rather than waiting for periodic major releases. This ongoing-standard model reflects how central and stable HTML has become: rather than being replaced by something newer, it keeps being extended.

HTML and the rise of AI-generated content

For essentially all of HTML's history, a person wrote it — by hand, with a code editor, or with a WYSIWYG tool. That's changed. AI assistants like Claude and ChatGPT can now generate complete, well-structured HTML documents directly from a plain-language description, drawing on the same enormous body of real-world HTML that shaped the language's practical, compatibility-first evolution through WHATWG. See why HTML has become AI's default output format for more on how that shift happened.

This new kind of authorship raises a genuinely new need: good tools for a human to inspect, review and refine HTML that a machine wrote. That's the same underlying idea that's guided HTML from the start — a structured, browser-native format anyone can read and use — now applied to a document a person didn't write themselves, but still needs to make their own. A visual HTML editor is one answer to exactly that need: open what the AI produced, see it rendered as the real page it is, and edit it directly, without needing to read a single tag.