Gnani AI launches Prisma v3.0 speech model for nine Indian languages plus English

Picture: Analytics India · see the original ↗

Gnani AI's new Prisma v3.0 speech recognition model recorded a macro Word Error Rate of 16.93 on the Vistaar benchmark, against 19.98 for ElevenLabs Scribe v2, 20.51 for Microsoft Azure Speech-to-Text and 26.57 for Deepgram Nova 3, per the company. The Bengaluru firm's model supports nine Indian languages plus English, handles mid-sentence code-switching, and offers four transcription modes, plus custom vocabulary prompting.

Source: Analytics India ↗