Sarvam AI Releases Saaras V4 Speech Model and Vision 2.1 for Indian Languages
Sarvam AI has launched Saaras V4, a multilingual speech recognition model supporting 22 Indian languages and English, capable of producing five transcript formats including verbatim, transliteration, and translation. The model handles code-mixed speech and regional accents with low latency, suitable for real-time applications. Additionally, Sarvam released Sarvam Vision 2.1, an upgraded vision-language AI model focused on document intelligence, improving accuracy in processing complex forms, tables, and handwritten Indic scripts across 22 languages while reducing operational costs.
First-hand measurement across 3 sources
We measured how 3 outlets covered this story. No outlet gave this story a measurable political slant — there is no left–right reading to report. Overall sentiment is positive (64/100). Lens Score 35/100.
Outlets measured: economictimes, businessstandard, mint. See how each one headlined and framed the same story in the source comparison below.
AI Analysis
Sentiment was consistent across outlets (50–75/100), indicating broadly factual reporting rather than editorialising.
Coverage timeline
mint broke this story on 25 Sept, 04:51 am. Other outlets followed.