Languages

Cleft is currently optimized for US English, using the Whisper model for voice recognition.

However, due to requests, we have enabled further languages to be available in early access. The Whisper model has been trained on 98 languages and we support 50+ languages that have reach a benchmark: The languages listed below have surpassed the industry standard benchmark of a < 50% word error rate (WER) for speech-to-text accuracy. We prefer to set expectations, rather than over-promise, so please be aware this will upgrade in the future, but this is not something we can guarantee today ✌️

For our multi-lingual users, it works like this:

• You select your desired transcription/input language or auto-detect language.

Some of the languages currently supported:

• Afrikaans

• Arabic

• Armenian

• Azerbaijani

• Belarusian

• Bosnian

• Bulgarian

• Catalan

• Chinese

• Croatian

• Czech

• Danish

• Dutch

• English

• Estonian

• Finnish

• French

• Galician

• German

• Greek

• Hebrew

• Hindi

• Hungarian

• Icelandic

• Indonesian

• Italian

• Japanese

• Kannada

• Kazakh

• Korean

• Latvian

• Lithuanian

• Macedonian

• Malay

• Marathi

• Maori

• Nepali

• Norwegian

• Persian

• Polish

• Portuguese

• Romanian

• Russian

• Serbian

• Slovak

• Slovenian

• Spanish

• Swahili

• Swedish

• Tagalog

• Tamil

• Thai

• Turkish

• Ukrainian

• Urdu

• Vietnamese

• Welsh