Norwegian audio and video to text

Dialects and the lack of a spoken standard

Norway has no single spoken standard, and people are proud of their dialects: Oslo and eastern speech, Bergen, Trøndelag, Stavanger and Northern Norway sound clearly different. Strong dialect words are the biggest source of errors, and a transcript of dialect speech is typically written in a standardised form that you may want to adjust.

Bokmål, Nynorsk and special letters

The two written standards are Bokmål and Nynorsk. This page uses one Norwegian setting, and there is no Nynorsk switch, so if you need Nynorsk spelling you will have to convert it in the editor. The letters æ, ø and å are kept as typed. Norwegians commonly write standard Bokmål even when speaking dialect.

What Norwegian speakers use it for

Student lecture recordings, interviews, meeting notes, podcasts, and subtitles for videos. A transcript of a dialect interview is often used as a working draft, so proof-reading against the recording matters. Pick Norwegian, not Swedish or Danish, even if the languages seem similar, because a wrong setting produces odd words. Keep names of people, farms and places in a list so you can search the finished text for them.

Which model works best for Norwegian

Norwegian is supported with a good amount of training material, but less than the largest languages, so the small model makes more slips than it does for English or Spanish, especially on rare words, names and fast or noisy speech. Expect a draft that needs a proper read-through.

On tryscriptorium.com the no-account preview and regular accounts run the small model (the larger medium and large-v3 models run when you self-host Scriptorium). In an account, pick Small in the model menu rather than Auto: Auto prefers the fastest, least accurate model for long files or a busy queue.

The small model copes with clear Bokmål-style speech but has a harder time with strong dialects, place names and people's names. Because Whisper treats Norwegian as one language, choosing it explicitly is more reliable than auto-detect, which can confuse neighbouring languages.

Subtitles and captions in Norwegian

Norwegian subtitles normally run two lines of up to about 37 to 42 characters. If the speaker uses a strong dialect, decide before exporting whether the subtitles should follow the speech or standard Bokmål, and edit consistently. Files are UTF-8, so æ, ø and å display everywhere.

The uploader is set to Norwegian

The upload box on this page transcribes as Norwegian, so the recogniser does not have to guess the language from the first seconds. If the recording is in another language or switches languages, use Audio to text, which detects the language automatically. Automatic transcription is a draft and no accuracy figure is promised: check names, numbers and anything you will rely on.