Rolong speech: transcription apps compared for Setswana; Setswana-English conversation was recorded in a church in the Gaborone study.
This page compares transcription apps and services for Setswana.
Speakers. About 8.2 million speakers in total: about 4.1 million native in South Africa (2011), 1.1 million native in Botswana (1993), 2.7 million second-language in South Africa (2002); Namibia and Zimbabwe numbers are unknown.
Writing. Latin. Tone is not marked, which the source says may lead to ambiguity. Digraphs include ph, th, tl, tlh, tsh, ng.
Tone is not marked, which can create ambiguity. The letter š was introduced in 1937, though 'sh' is still sometimes written, so both spellings occur in transcripts. ê and ô appear in textbooks and reference works but are less common in everyday writing. Digraphs are ph, th, tl, tlh, tsh and ng.
| Option | What its page shows about Setswana | Source |
|---|---|---|
| Autrans | Autrans offers Setswana. Measured result: Word error rate 62.2% on read speech, NCHLT test set, measured Sep 2026. | autrans.online |
| Sonix | The list on the Sonix page does not include Setswana. | sonix.ai |
| Descript | The list on the Descript page does not include Setswana. | help.descript.com |
| Trint | The list on the Trint page does not include Setswana. | info.trint.com |
| Maestra | The list on the Maestra page does not include Setswana. | maestra.ai |
| VEED | VEED names Setswana on its page. It says: "Transcription: Afrikaans, Amharic, Arabic, Hausa, Igbo, Lingala, Shona, Somali, Swahili, Tsonga, Tswana, Xhosa, Yoruba, Zulu." | support.veed.io |
| TurboScribe | The pages read for TurboScribe say nothing about Setswana. | turboscribe.ai |
| Rev | The pages read for Rev say nothing about Setswana. | rev.com |
| Happy Scribe | The pages read for Happy Scribe say nothing about Setswana. It says: "Yoruba, Hausa, Igbo, Shona, Somali, Xhosa, Wolof, Lingala, Twi, Oromo, Setswana and Nigerian Pidgin are not in the table." | happyscribe.com |
Named: its page names the language. Not listed: its own list leaves the language out. Not stated: the pages read say nothing either way, so test a file. All pages read on 22 October 2026.
| API | What its page shows about Setswana | Source |
|---|---|---|
| Speechmatics | The list on the Speechmatics page does not include Setswana. | docs.speechmatics.com |
| Intron | Intron names Setswana on its page. | docs.voice.intron.io |
| Lelapa AI | The list on the Lelapa AI page does not include Setswana. It says: "Xhosa, Tswana, Tsonga, Swahili, Shona, Hausa and Yoruba appear only for text translation, so they are 'no' for speech-to-text." | docs.lelapa.ai |
| AssemblyAI | The pages read for AssemblyAI say nothing about Setswana. | assemblyai.com |
| Spitch | The pages read for Spitch say nothing about Setswana. | docs.spitch.app |
These are APIs for developers, not apps for a recording.
Word error rate 62.2% (95% confidence interval 57.8–66.6%) 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 turns audio, video, WhatsApp voice notes and phone call recordings into a transcript you can read, search, edit and export.
Languages here: Setswana.
Yes, Setswana is one of the live languages. Word error rate 62.2% (95% confidence interval 57.8–66.6%) on read speech, NCHLT test set, measured Sep 2026.
VEED names Setswana on its own pages, as read on 22 October 2026. Sonix, Descript and Trint do not list it.
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.
Last checked 22 October 2026. How we measure accuracy. Spotted a mistake? Write to hello@autrans.online.