Language Detector | Identify Text Languages Online

Compare text of at least 10 characters with franc n-gram statistics in the browser and view the nearest language candidates with relative scores.

Focus mode

Text

Identify the language as you type, with no downloads or server uploads.

Example text

Result

Statistical n-gram language detection (franc).

Candidate scores are relative values, not calibrated probabilities or confidence levels.

Korean

한국어 · kor

Candidates (relative scores)

Korean

100

Key features

Compares text across a curated set of 30 named languages
Shows the top language's English name, native name, and ISO 639-3 code
Lists up to five candidates with relative scores
Runs with bundled franc statistics and no separate AI model download

How to use

Enter or paste at least 10 characters of text.
Review the leading language and the ranked candidate list.

Examples

Triage an unlabeled passage
Paste a substantial excerpt before translating or reviewing it to see which language candidates rank first.
Inspect mixed-language copy
Compare the ranked candidates for blended text while treating the score as relative evidence, not a final answer.

Privacy

Your text is analyzed in the browser with bundled franc statistics and is not sent to a server or model provider for detection.

Requirements

Short passages, mixed-language text, and text dominated by names can produce unreliable candidates. Scores compare candidates and are not calibrated probabilities.

Common questions

Does a high score prove the language?: No. The score ranks candidates against one another; it is not a calibrated probability or certainty.
Where is my text analyzed?: Detection uses bundled franc statistics in your browser, without sending text to a server or model provider.