India showcased three homegrown artificial intelligence models at the India AI Impact Summit 2026 in New Delhi, signalling a strong push towards building sovereign AI systems powered by domestic data, languages and computing infrastructure.
The announcements reflect a broader strategy to reduce reliance on global platforms and develop AI technologies tailored to India’s linguistic diversity and governance needs. Officials said the focus is on creating scalable solutions for sectors such as public services, education, healthcare and enterprise applications.
Sarvam AI Introduces Two Indigenous Large Language Models
Indian startup Sarvam AI launched two large language models — a 30-billion-parameter model designed for real-time conversational use and a 105-billion-parameter model capable of handling complex reasoning and long-form tasks. The models support all 22 scheduled Indian languages and are optimised for voice-first interactions, aiming to expand access among non-English speakers.
The smaller model offers a 32,000-token context window for efficient deployment, while the larger model supports up to 128,000 tokens and uses a mixture-of-experts architecture to balance performance and cost. The company also demonstrated a multilingual chatbot named “Vikram,” highlighting conversational AI tailored for Indian users.
Gnani.ai Launches Multilingual Voice AI System
Bengaluru-based Gnani.ai introduced Vachana TTS, a text-to-speech system capable of cloning voices using less than 10 seconds of recorded audio. The system can generate speech in 12 Indian languages while preserving tone and style, and is designed for low-bandwidth environments such as government services and customer support platforms. The company said all data processing is hosted within India to ensure data sovereignty.
BharatGen Unveils Param2 Multilingual Foundation Model
Under India’s sovereign AI initiative BharatGen launched Param2, a 17-billion-parameter multilingual foundational model intended for governance, agriculture, healthcare and educational applications. The model was developed in collaboration with NVIDIA using advanced AI infrastructure to support scalability.
Also Read | Netherlands PM Dick Schoof Praises India’s AI Momentum at India AI Impact Summit 2026
Focus on AI Sovereignty and Local Innovation
Officials said the launches demonstrate India’s move toward AI self-reliance, emphasising local control over datasets, computing resources and model development. By prioritising Indic languages and voice-driven interfaces, the initiative aims to make AI tools accessible to a broader population while addressing real-world challenges.
The new models are expected to strengthen India’s domestic AI ecosystem, encourage innovation among startups and research institutions, and support the country’s long-term vision of building inclusive and locally relevant digital technologies.