India AI Impact Summit 2026: Gnani.ai launches 5-billion-parameter voice AI model under IndiaAI Mission
India AI Impact Summit 2026: The system, called Inya VoiceOS, was inaugurated by Prime Minister Narendra Modi at the India AI Impact Summit 2026 and released as a research preview ahead of a larger model planned by the company
- Feb 17, 2026,
- Updated Feb 17, 2026 10:06 AM IST
India AI Impact Summit 2026 | Homegrown artificial intelligence (AI) firm Gnani.ai on 17 February unveiled a 5-billion-parameter voice-to-voice foundational model designed to enable natural spoken interaction across multiple Indian languages, as the country pushes to build sovereign AI capabilities.
The system, called Inya VoiceOS, was inaugurated by Prime Minister Narendra Modi at the India AI Impact Summit 2026 and released as a research preview ahead of a larger model planned by the company.
Unlike conventional speech systems that convert speech to text and back again, the model processes audio directly, allowing more natural conversations and faster responses, the company 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,” Ganesh Gopalan, co-founder and chief executive of Gnani.ai, said.
“With Inya VoiceOS, we're bringing emotion and context into every interaction while delivering significantly better accuracy than traditional cascaded systems,” he added.
The Bengaluru-based company said the model was trained on more than 14 million hours of multilingual speech data and over 8 trillion text tokens, with support for more than 15 Indian languages.
Inya VoiceOS operates directly in acoustic and semantic space, enabling it to capture tone, emotion and pacing while handling interruptions and overlapping speech without restarting conversations, according to the company.
The company said the technology could be used for government helplines, emergency response systems and enterprise applications such as banking, healthcare and logistics.
Gnani.ai added that the model was built, trained and deployed entirely within India, positioning it as part of the country’s broader effort to ensure data sovereignty in artificial intelligence.
The company said it plans to follow the research preview with a larger 14-billion-parameter voice-to-voice model.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
India AI Impact Summit 2026 | Homegrown artificial intelligence (AI) firm Gnani.ai on 17 February unveiled a 5-billion-parameter voice-to-voice foundational model designed to enable natural spoken interaction across multiple Indian languages, as the country pushes to build sovereign AI capabilities.
The system, called Inya VoiceOS, was inaugurated by Prime Minister Narendra Modi at the India AI Impact Summit 2026 and released as a research preview ahead of a larger model planned by the company.
Unlike conventional speech systems that convert speech to text and back again, the model processes audio directly, allowing more natural conversations and faster responses, the company 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,” Ganesh Gopalan, co-founder and chief executive of Gnani.ai, said.
“With Inya VoiceOS, we're bringing emotion and context into every interaction while delivering significantly better accuracy than traditional cascaded systems,” he added.
The Bengaluru-based company said the model was trained on more than 14 million hours of multilingual speech data and over 8 trillion text tokens, with support for more than 15 Indian languages.
Inya VoiceOS operates directly in acoustic and semantic space, enabling it to capture tone, emotion and pacing while handling interruptions and overlapping speech without restarting conversations, according to the company.
The company said the technology could be used for government helplines, emergency response systems and enterprise applications such as banking, healthcare and logistics.
Gnani.ai added that the model was built, trained and deployed entirely within India, positioning it as part of the country’s broader effort to ensure data sovereignty in artificial intelligence.
The company said it plans to follow the research preview with a larger 14-billion-parameter voice-to-voice model.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine