Fact · Languages and accents
34.6BLEU points
On the Fleurs read-speech test, Whisper large-v2 scored 34.6 BLEU when translating German speech into English.
Checked against its source on October 17, 2026
In the source's own words
“Whisper large-v2 34.6 23.7 80.2 23.3 18.7 19.6 22.1 24.4 32.2 27.9 16.2 0.4 21.8 22.0”
Robust Speech Recognition via Large-Scale Weak Supervision (full text), Alec Radford et al., OpenAI (arXiv), 2022, Table 14, BLEU scores on Fleurs (speech translation).
Open the original (DOI 10.48550/arXiv.2212.04356)- What it measures
- The table has one column per language. The paper uses the English transcript of each sentence as the reference translation, and the German value is the first column of the Whisper large-v2 row.
- What it does not show
- BLEU counts word overlap with one reference translation, so it is only a rough guide to meaning. The test uses read sentences in a quiet setting, not conversations or noisy recordings.