use case · basics
The simplest thing NER does: read text and colour the named things — people, organizations, locations, and other proper nouns — with a confidence for each. Pick an example or type your own and watch the highlights update as you type.
The model was trained on news text (CoNLL-2003), so it's sharpest on public figures, companies and places, and hazier on rare or invented names. Hover any highlight to see how confident it was — always a better signal than the label alone.
← Back to BERT NER · Next: Practical →