Fact · Languages and accents
A 2024 Urdu benchmark found that Seamless-large was best on read speech, while Whisper-large was best on conversational speech.
Checked against its source on October 17, 2026
In the source's own words
“seamless-large outperforms other ASR models on the read speech dataset, while whisper-large performs best on the conversational speech dataset”
WER We Stand: Benchmarking Urdu ASR Models, Samee Arif et al., arXiv, 2024, Abstract.
Open the original- What it measures
- The benchmark tests Whisper, MMS and Seamless-M4T on Urdu, and adds the first conversational dataset for Urdu.
- What it does not show
- Models are ranked on two datasets written by the authors. Results on one type of speech do not carry over to the other.