During the ongoing India AI Impact Summit 2026, sovereign artificial intelligence models developed in India found special mentions as minister of electronics and information technology Ashwini Vaishnaw and also Prime Minister Narendra Modi hailed national sovereignty when it comes to artificial intelligence.
Vaishnaw said that India’s “bouquet of sovereign models”, many of which were launched at the summit itself, are “focused on multi-modal and multilingual capability”.
While no exclusive AI models developed in India were mentioned in PM’s speech, it did put in focus India’s three sovereign AI models — Sarvam AI, Gnani.ai, and BharatGen.
Track live updates of India AI Impact Summit 2026 here.
SarvamAI
SarvamAI, a Bengaluru-based startup, describes itself as “Sovereign by design”, adding “Build, deploy, and run AI with full control, developed and operated entirely in India”.
“We want India to embrace the most important technological shift of our time with confidence and control. Our ambition is to build foundational components and apply them to the country's unique needs. To this end, we've built a full-stack AI platform, with everything developed, deployed, and governed entirely in India,” the company’s website says.
Also read: ‘Right direction, democratised’: PM Modi explains how AI can benefit all
{{/usCountry}}Also read: ‘Right direction, democratised’: PM Modi explains how AI can benefit all
{{/usCountry}}During the ongoing AI Summit, Sarvam AI unveiled an artificial-intelligence model that is tailored more for Indian needs as compared to its global counterparts such as ChatGPT and Claude.
It announced two models at the summit which are built to be used through voice commands and are accessible through 22 Indian languages. The company says that this would give them an edge over their competitors in India with a population of 1.45 billion people where a vast majority of people are not comfortable with English, reported Bloomberg.
“Today we show we can bring our own AI to a billion Indians,” Sarvam co-founder Pratyush Kumar was quoted as saying.
“Sovereignty matters much more in AI than building the biggest models,” he added.
Also read: Bill Gates a no-show at India AI summit, foundation issues statement
Gnani.ai
In India’s sovereign AI push, Gnani.ai launched Vachana TTS voice AI at the ongoing AI Impact Summit, which clones voices in 12 Indian languages using audio which is 10 seconds long or lesser.
The company describes the feature as “Voice models that truly understand India” available in “11 Indian languages. Every accent. Any language mix.”
It said that the model is engineered for India's real voices — “accurate through accents, language switches, and natural, everyday speech”.
Gnani.ai formally introduced Vachana STT during the ongoing summit as part of IndiaAI Mission and described it as a foundational, enterprise-grade Indic speech-to-text model trained on over one million hours of real-world voice data.
BharatGen
BharatGen, which describes itself as India’s first sovereign AI which is “by Bharat, for Bharat”, says it is “building an ecosystem of partners—startups, system integrators, government agencies, and VCs—to co-create AI applications for good governance, defense, agriculture, and commercial use.”
During the AI Impact Summit, BharatGen unveiled the Param2 17B MoE, a 17-billion-parameter multilingual Mixture-of-Experts (MoE) foundational modeI.
Param2 17B’s training pipeline was developed in collaboration with NVIDIA, and according to the firm, its model was as part of its larger goal to build India-first sovereign AI systems that capture the country’s linguistic and cultural diversity, reported Money Control.
“We are extremely delighted at unveiling PARAM 2 17B at the AI Impact Summit in Bharat Mandapam and also at the partnership with NVIDIA. As a truly sovereign AI initiative, the launch and unveiling of PARAM 2 17B marks a pivotal moment for India and we at BharatGen are elated at contributing to the country’s AI progress,” said BharatGen’s CEO Rishi Bal, as quoted by the aforementioned report.