Gnani.ai launches Inya VoiceOS, India’s first 5B voice-to-voice model

Gnani.ai today announced the deployment of Inya VoiceOS, India's first 5-Billion-parameter voice-to-voice foundational AI model, built under the India AI Mission. Inya VoiceOS enables end-to-end spoken intelligence by operating directly in acoustic and semantic space, eliminating the need for intermediate speech-to-text and text-to-speech pipelines. Gnani.ai joins a select group of companies worldwide that have built this advanced architectural capability.

Inya VoiceOS is trained on one of the largest sovereign voice datasets for Indian languages, comprising 5 billion parameters, 14+ million hours of multilingual speech pretraining data, 1.2+ million hours of task-specific speech finetuning data, and 8+ trillion text tokens for linguistic grounding and reasoning. 

Unlike traditional cascaded systems, Inya VoiceOS, consumes and generates speech tokens directly and jointly encodes phonetics, prosody, semantics, and intent, while preserving paralinguistic cues such as tone, emotion, pacing, and pauses. The model supports streaming, interruption-aware inference, handles overlapping speech and mid-utterance corrections without forcing conversational resets.

In terms of technical performance, the model has sub-second end-to-end latency, 24 kHz audio output with natural prosody, native support for 15+ Indian languages and a robust handling of code-mixed speech.

Where can this model be used?

The statement highlighted the following usage:

Government Services: Conversational AI for helplines, grievance redressal, and emergency response systems enables natural, multilingual citizen interaction with emotion-aware routing.

Enterprise Applications: Voice-driven workflows across BFSI, healthcare, insurance, and logistics support hands-free operation and context-aware conversational automation.

The model was inaugurated by Prime Minister Narendra Modi at the India AI Impact Summit 2026, with Co-Founder and CEO Ganesh Gopalan.

Speaking at the event, Gopalan said, "Voice-to-voice AI is not just about faster pipelines. It is about a fundamentally different architecture that preserves what makes human conversation effective. With Inya VoiceOS, we're bringing emotion and context into every interaction while delivering significantly better accuracy than traditional cascaded systems. This is critical for building AI that truly understands and responds to the nuances of human speech."

Gnani.ai is an India-born agentic AI and voice Infrastructure company building foundational voice models and systems for enterprises and governments. Fluent in over 15 Indian languages, Gnani.ai works with 200+ large organisations, including Tata Group, Mahindra Group, and Air India, to deploy speech-first AI at population scale.