Fact · Speech recognition
The turbo Whisper model is not trained for translation tasks, so translating non-English speech into English needs a multilingual model such as medium or large.
Checked against its source on October 17, 2026
In the source's own words
“However, **the `turbo` model is not trained for translation tasks**.”
openai/whisper: README (raw text), OpenAI, GitHub (openai/whisper), 2026, Command-line usage.
Open the original- What it measures
- The README adds that you should use the multilingual models tiny, base, small, medium or large to translate, and that turbo returns the original language even when the translate task is set.
- What it does not show
- The README gives this as guidance for the released models. Check translation quality on your own audio, because the README gives no accuracy figures for translation.