Fact · Speech recognition
119hours
Earnings-22 is a corpus of 125 files totaling 119 hours of English earnings calls from companies around the world.
Checked against its source on October 17, 2026
In the source's own words
“we present Earnings-22, a 125 file, 119 hour corpus of English-language earnings calls gathered from global companies”
Earnings-22: A Practical Benchmark for Accents in the Wild, Miguel Del Rio et al., arXiv (Computation and Language), submitted to Interspeech 2022, 2022, Abstract.
Open the original- What it measures
- Earnings-22 was released as a free benchmark for accented English in real recordings.
- What it does not show
- The corpus is English only. It is spontaneous, financial speech, so it is one genre among many.