Kelvin Sign K vs Letter K: Why Unicode Treats Them as Equivalent

The Kelvin sign K is a Unicode relic you may never notice visually. It usually looks exactly like ordinary capital K, and Unicode deliberately treats the two as canonically equivalent. Modern kelvin notation uses the ordinary letter.

So for current notation, write kelvin as K. Choosing K does not make the value more scientific; it only introduces a legacy code point.

K and K compared

Why K exists at all

Unicode inherited characters from earlier encoding standards, including a dedicated Kelvin sign. Keeping U+212A made it possible to convert those older documents without losing which code point the source system had used.

Unicode does not preserve K as a separate modern spelling of kelvin. U+212A and ordinary K are canonically equivalent, and the Core Specification recommends the regular letter for text you author today.

Why normalization turns K into K

Because the equivalence is canonical, normalization can replace K with K even under forms such as NFC. A workflow that expects U+212A to remain distinct after Unicode normalization is therefore relying on a distinction the standard intentionally removes.

If exact historical encoding matters, preserve the original source separately. For search, indexing, APIs, and newly authored scientific text, ordinary K is the stable choice.

Kelvin’s actual SI symbol is K

Kelvin is one of the SI base units, and its symbol is the ordinary capital K. A value such as 300 K follows the usual spacing seen across measurement notation: number, space, unit. No kelvin-specific Unicode character is required.

The symbol stays K regardless of the numerical value. Plurality belongs to the written unit name in prose, not to the symbol, so a table shows 1 K, 20 K, and so on.

Kelvin is not written with a degree sign in current SI notation. Expressions such as 300 °K reflect historical usage rather than the present symbol. The difference between °C and K is covered in Celsius °C vs Kelvin K.

Why an ordinary letter can serve as a unit symbol

At the Unicode level, U+004B is simply LATIN CAPITAL LETTER K. Scientific convention gives that character the role of the kelvin symbol when it appears in a quantity value. The code point does not need a special unit-specific name.

The same idea appears throughout SI notation: ordinary letters are used as symbols for metre, second, ampere, mole, and many other units or prefixes. Meaning comes from the technical convention and surrounding expression, not from selecting a decorative Unicode variant.

Lowercase k is a different symbol

The lowercase form has another job entirely: k is the kilo prefix, 10³. A function that changes letter case can therefore alter the measurement token rather than merely restyling it.

For example, 300 K is a temperature. In 10 kΩ, the lowercase k multiplies the ohm by one thousand. The dedicated k vs K guide focuses on that distinction.

Where K still survives

U+212A still appears in old databases, copied PDFs, OCR output, historical standards, and character palettes. A spreadsheet may visibly show 300 K while the copied cell actually contains U+212A; normalization then turns it into the ordinary letter without changing the temperature. Its presence says something about the source encoding, not about a different scale.

Seeing K become K during normalization is expected. An archive that cares about the original code point can keep the raw source alongside a normalized searchable value instead of making one field serve both purposes.

For ordinary modern text, K adds no useful distinction. Before normalization, it can simply make two visually identical temperature values look different to software.

The same Unicode history appears in OHM SIGN Ω vs Ω and ANGSTROM SIGN Å vs Å, where regular characters are likewise preferred for new text.

Not every K means kelvin

The symbol K identifies the unit kelvin when it follows a numerical value in the appropriate context. It should not be expanded mechanically every time the letter K appears. In ordinary prose, product codes, variables, or names, U+004B remains simply a capital Latin letter.

This matters to automated tagging. A rule that converts every standalone K into a “kelvin entity” will create false positives, just as a rule that converts every m into metre would. Unit recognition should consider surrounding numbers, expressions, and domain context.

Use ordinary K for current text

Use ordinary K U+004B for kelvin. Keep K only when reproducing legacy text in which the original code point is itself part of what you need to preserve.

When processing data, normalize before comparison and keep generic case conversion away from unit strings. If the source code point has archival value, preserve it separately while using K for the current notation readers should see.