Kencorpus Swahili ASR

What it is

Kencorpus is a Kenyan academic-led initiative that published a multilingual Swahili-English-Dholuo speech and text corpus, along with baseline ASR models. The resource is foundational for any East-African transcription project — particularly those that need to handle Swahili-English code-switching, which is the dominant register in Nairobi and other urban markets. Best fit when the buyer is swahili and swahili-english code-switching transcription in east african contexts. The honest caveat: primarily a dataset; productisation is the integrator's responsibility. As with any open-weights release, the integrator owns hosting, scaling, and SLA — but the licensing cost is zero and the model can be fine-tuned on in-house audio.