Font Comparison
Set the same text in up to four fonts at once, under identical size, weight, tracking and leading – with the x-height and cap-height that explain why two fonts at the same size can look nothing alike.
Drop fonts here, or click to choose
Up to four · .ttf .otf .woff .woff2 · 0 of 4 loaded
Every font is loaded into this tab only. Nothing is sent to this site, and nothing is stored.
How to use this tool
- Drop two to four font files onto the box above. You can select several at once.
- Type the text that matters – the real headline or the real product name, not a pangram, if the decision is about that specific word.
- Set size, weight, tracking and leading once. Every column follows them, so any difference you see is a difference in the font.
- Compare the metrics rows underneath. They usually explain what your eye is already telling you.
- Stack vertically on a narrow screen, or when you want to compare long lines rather than short ones.
Why "Hamburgefonstiv"
It is the traditional type-design test word, and it is not arbitrary. It contains the shapes that carry most of a typeface's character: a round o, a straight-sided n, the join of an a, ascenders, descenders, and the awkward g. Judge those and you have judged most of the font.
Replace it with your own words as soon as the decision is about specific text. A font that handles a pangram beautifully can still make your client's surname look wrong.
Why two fonts at the same size look like different sizes
Because "16px" describes the em box, not the letters. What your eye reads as size is mostly x-height – the height of a lowercase x as a proportion of the em. A face with a 52% x-height set at 16px looks noticeably larger than one with a 45% x-height at the same 16px.
That is why each column shows x-height and cap-height as a percentage of the em rather than in raw units: 1062 and 1493 mean nothing side by side, but 52% and 73% are directly comparable. If you are replacing one font with another, matching those percentages matters more than matching the pixel size, and ignoring them is how a redesign ends up looking subtly wrong everywhere.
A note on the weight selector
If a font in a column has no real file for the weight you pick, the browser synthesises it — and it will do that silently, per column, which can make a comparison misleading. When you are comparing seriously, load the actual weight file for each candidate rather than asking for 700 from four regulars.
Why the metrics row sometimes shows dashes
WOFF2 files are compressed in a way the metrics parser on this page cannot read directly, and loading 315 KB of decompressor to fill in four numbers is not a fair trade for everyone else's page weight. The comparison itself works perfectly with WOFF2 – only the metrics row is blank. Drop the TTF or OTF instead if you need those figures, or read them from the font analyzer, which does carry the decompressor.
To examine one font in depth rather than several at once, the font previewer shows it as a heading, paragraph, button and nav. To check whether a candidate covers the languages you need, use the language checker.