Fact · Languages and accents
The Whisper README says that the turbo model returns the original language when the translate task is set, and recommends the medium or large model for the best translation results.
Checked against its source on October 17, 2026
In the source's own words
“Use `medium` or `large` for the best translation results.”
openai/whisper: README (raw text), OpenAI, GitHub (openai/whisper), 2026, Translation section of the README, after the translate example.
Open the original- What it measures
- The README's note follows its example of translating Japanese speech into English with the medium model and the translate task.
- What it does not show
- The README gives no accuracy figures for this recommendation. It is the project's advice, not a measured comparison of the two models.