Sarvam AI launches Saaras V4 speech recognition model for 22 Indian languages

Sarvam AI has introduced the Saaras V4 automatic speech recognition (ASR) model, which supports five output modes and 22 Indian languages. The model is designed for multilingual voice applications and features low-latency streaming capabilities. This launch marks an advancement in speech recognition technology tailored for diverse Indian linguistic needs.

Originally published by gnews