Autrans understands Sesotho (Southern Sotho), a national language of Lesotho and South Africa, the way it is actually spoken.
Sesotho is spoken across Lesotho and South Africa. Autrans transcribes Sesotho keeping its disjunctive spelling, concords and tense markers written as separate words, and the special letter š, without normalizing it into English. Visitors in Lesotho have Sesotho switched on by default. In South Africa, where Setswana is the default of the two, add Sesotho under Settings, Language and it swaps in for Setswana; anyone else can add it there too.
Word error rate 69.4% (95% confidence interval 65.7–73.6%) on read speech, NCHLT test set, measured Sep 2026.
Measured on 250 recordings. Word error rate is the share of words that differ from a human-written transcript, so lower is better. Read speech means people reading set sentences aloud, so conversation, background noise and phone-quality audio will usually score worse.
Test set: NCHLT speech corpus (SADiLaR).
Autrans handles standard Sesotho. Lesotho and Free State varieties differ slightly, and output reflects what the speaker says.
Visitors in Lesotho have it switched on by default. In South Africa the default is Setswana, so add Sesotho under Settings, Language and it swaps in for Setswana; anyone else can add it there too.
No, and that is deliberate. The two are close enough that offering both made the language labels wrong for roughly one segment in six in our tests, even though the words were heard correctly. Choose the one your recordings are in; adding one swaps the other out of your list.
The measured figure above is the one to go by. It comes from people reading set sentences aloud, so conversation and noisy audio will usually score worse. Clear audio gives the best results; review the transcript before relying on it for anything critical.
Send Autrans into your next meeting, use it in the room, or upload a recording. Your first 30 minutes are free, given once, and no credit card is needed to start.