Abridge uses proprietary automatic speech recognition (ASR) to process audio, and our AI performs well across diverse accents, languages, and dialects.
Abridge supports 28 official languages with state-of-the-art ASR accuracy, outputting a fully English note at the end:
English
Spanish
Cantonese Chinese
Mandarin Chinese
Tagalog
Vietnamese
Arabic
French
Korean
Russian
Haitian
Creole
German
Japanese
Hindi
Polish
Romanian
Finnish
Turkish
Hungarian
Bulgarian
Catalan
Danish
Czech
Portuguese
Indonesian
Ukrainian
Dutch
What are high performing languages?
High-performing languages are languages our machine learning science team has dedicated to significant time and resources to evaluate, benchmark, and optimize our models against. As a result, we perform ~1% absolute (Word Error Rate) better than the leading state-of-the-art models on these languages. In the event that the audio has been mis-transcribed, the ground-truth audio is available for you to refer to and ensure that your note captures all critical details.