Occitan speech-to-text
Not usableWe sent 30 Occitan test clips through this API. Whisper Large v3 made 71% word errors, which is not good enough to read as a transcript. This page shows what came back so you can judge for yourself.
Measured accuracy
| Model | Word error rate | Character error rate | Clips with no errors | Empty transcripts | Speed |
|---|---|---|---|---|---|
Whisper Large v3openai/whisper-large-v3 | 71% | 23% | 0 of 30 | 0 of 30 | 0.038× real time |
Not usable. Most words come out wrong. Do not build on it for this language. The clips are clean, single-speaker read speech, so treat this as a best case for your own recordings. How we measured, and all 83 languages.
What a transcript looks like
The clip with the median error rate from the test, reference text against what the API returned.
| Spoken | Lo coeficient d'aspècte d'aqueste format (en devesissent per dotze per obtenir lo rapòrt de nombres entièrs mai simple) es doncas dich d'èsser de 3:2. |
| Whisper Large v3 | Le coefficient d'aspect d'acquest format en dévississant pè d'autres pèr obtenir le rapport de nombre entier mais simple. Est donc à 17, 2, 3, 2. |
Language detection
Send language=oc. Of 10 clips sent without it, only 0 were identified as Occitan; the others were labelled French (3), Lingala (3), Haitian Creole (2), English (2). When detection is wrong, Whisper writes the transcript in the language it guessed.
What to do instead
We would rather you know this before you build than after. For Occitan, look at services that train specifically on the language, such as Google Cloud Speech-to-Text or Azure AI Speech, both of which publish their language lists, or a Whisper checkpoint fine-tuned on Occitan from Hugging Face run on your own hardware. If your audio mixes Occitan with a well-supported language, the well-supported parts will still transcribe. You can confirm all of this on your own audio with the free credit on a new account: transcription playground.
FAQ
How accurate is Whisper for Occitan?
On 30 clean read-speech clips it made 71% word errors. Most words come out wrong. Do not build on it for this language. Real recordings with noise, accents or several speakers will score worse.
Does language auto-detection work for Occitan?
Not reliably. Of 10 clips sent without a language, 0 were identified as Occitan; the rest were labelled French (3), Lingala (3), Haitian Creole (2), English (2). A wrong label means the transcript is written in the wrong language, so always send language=oc.
Can it translate Occitan speech to English?
The API returns Occitan text. For English, transcribe first and translate the text with a chat model on the same key; with accuracy this low the translation will inherit the errors.