Aqua currently supports these languages for dictation:
Yes, Aqua can automatically switch between all supported languages. Set your language to “Auto-Detect” in settings.
Aqua uses its own proprietary inference engine and models to deliver state-of-the-art accuracy across all supported languages.
Yes, accuracy can vary between languages. Like most speech recognition systems, language accuracy correlates with the amount and quality of training data available for each language.
The accuracy distribution across our supported languages is similar to the word error rate patterns found in leading models like Whisper and Voxel, with widely-spoken languages typically showing the highest accuracy rates.
Aqua was specifically designed to adapt to:
Our models are trained to handle diverse accents and speaking styles, making Aqua particularly effective for users with non-native accents or those working in challenging audio conditions.
To use automatic language detection, simply set your language preference to "Auto-Detect" in Aqua settings.
Once enabled, each transcription session will automatically identify and transcribe the spoken language without requiring manual switching.
