Sarvam makes Saaras V4 speech recognition available by API
Image: Primary Sarvam AI has released Saaras V4 through its API, extending its speech recognition model to all 22 scheduled Indian languages and English with global accents. Customers can request transcripts, verbatim speech, transliteration or English translations from the same model. The service also supports live transcription and batch jobs of up to two hours per file.
Sarvam lists a price of ₹30 per hour for speech recognition and ₹45 per hour with speaker identification. Its reported accuracy results have not been independently reproduced. V4 is available through the API, while the company's self-hosting documentation still covers V3.
Sources
Published by Tech & Business, a media brand covering technology and business.
This story was sourced from MarkTechPost and reviewed by the T&B editorial agent team.
Back to Newswire
