article_page.html in HTML
An article page with social metadata, semantic structure and a figure.
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Why tokens per minute - typestar</title>
<link rel="canonical" href="https://typestar.io/blog/tokens">
<meta property="og:title" content="Why tokens per minute">
<meta property="og:image" content="https://typestar.io/og/tokens.png">
<meta property="og:type" content="article">
<link rel="stylesheet" href="/app.css">
</head>
<body>
<header>
<a href="/">typestar</a>
</header>
<nav aria-label="Breadcrumb">
<ol>
<li><a href="/">Home</a></li>
<li><a href="/blog">Blog</a></li>
<li aria-current="page">Why tokens per minute</li>
</ol>
</nav>
<main>
<article>
<h1>Why tokens per minute</h1>
<p>
By Rowan Vale, <time datetime="2026-07-29">29 July 2026</time>
</p>
<section>
<h2>Words are the wrong unit</h2>
<p>
Prose typing tests divide characters by five and call the
result a word. Code is mostly punctuation, so that number
says very little about how fast you can write a function.
</p>
</section>
<section>
<h2>Counting tokens instead</h2>
<p>
A lexer already knows where each token ends. Count the
tokens you finished and you measure the thing you care
about: how quickly ideas reach the file.
</p>
<figure>
<img src="/img/tpm.png" alt="Tokens per minute over six weeks"
width="720" height="360">
<figcaption>Six weeks of daily practice.</figcaption>
</figure>
</section>
<aside>
<h2>Try it</h2>
<p><a href="/tours/python/basics">Python basics tour</a></p>
</aside>
<footer>
<p>Filed under measurement</p>
</footer>
</article>
</main>
<footer>
<p>Built in the open.</p>
</footer>
</body>
</html>
How it works
- The head carries canonical and Open Graph tags for sharing.
- The article nests sections, a figure, an aside and its own footer.
timegives the publication date in a machine-readable form.
The run, in numbers
- Lines
- 71
- Characters to type
- 1642
- Tokens
- 359
- Three-star pace
- 80 tpm
At the three-star pace of 80 tokens a minute, this run takes about 269 seconds.
Step 1 of 1 in Encore, step 23 of 23 in Semantics & metadata.