← All models · POS tagging

Basics · token-classification · WASM

Colour every word by its part of speech

The fundamental capability, felt directly: the model reads a French sentence and gives every word one grammatical tag. Edit the text and watch the colours shift — a noun becomes a verb, an adjective appears. Hover any word for its raw French TreeBank tag and confidence.

Tag a sentence

“The little boy throws the red ball in the garden.”

Keep going

Practical →

Extract keywords by POS and profile a text's writing style.

Wild →

Mad-Libs: blank words by part of speech and fill the gaps.

Multi-model →

POS + fill-mask reimagine the words into a new sentence.