Fact · Transcription
The authors state that Whisper's training data includes no isiXhosa and no closely related languages, so they chose Swahili as the language token when they fine-tuned on isiXhosa.
Checked against its source on October 17, 2026
In the source's own words
“On the other hand, no isiXhosa or any closely related languages are included in Whisper.”
Speech Recognition for Automatically Assessing Afrikaans and isiXhosa Preschool Oral Narratives, Christiaan Jacobs et al., arXiv (Stellenbosch University and University of Edinburgh), 2025, Section III, base speech recognition model.
Open the original- What it measures
- The authors say Whisper was trained on 680k hours of speech in 97 languages, including 4.1 hours of Afrikaans and thousands of hours of related Germanic languages such as Dutch and German. They chose Swahili as the nearest available language token for isiXhosa.
Source
Cited on
- Dictation Languages: What Each App and Device SupportsGuide
- How to Translate Audio and Video to English TextGuide
- Can AI transcribe Zulu, Xhosa and other African languages?Question
- Can you translate Zulu or Xhosa audio to English?Question
- Is speech recognition less accurate in languages other than English?Question