Sovereign AI leap at India AI Impact Summit: 12 indigenous foundation models power multilingual, inclusive innovation

New Delhi [India], February 14 (ANI): The India AI Impact Summit will be held from February 16 to 20 at Bharat Mandapam in the national capital, bringing together policymakers, technology experts, researchers, startups, and entrepreneurs from across the country and abroad to discuss the evolving landscape of artificial intelligence.
The summit is being convened by the Ministry of Electronics and Information Technology (MeitY) under the IndiaAI Mission.
The event is envisioned as a significant international forum to frame a forward-looking roadmap for inclusive, responsible and impactful AI, with an emphasis on generating concrete outcomes that contribute to economic advancement, social progress and the sustainable adoption of AI technologies.
Anchoring these ambitions, the five-day deliberations will be structured around three core pillars, "People, Planet, and Progress."
Conversations under these themes will focus on employment generation and skilling, environmentally sustainable and energy-efficient AI systems, and the contribution of AI to economic and social development.
Seven thematic working groups, co-led by representatives of the Global North and Global South, will put forward concrete recommendations covering initiatives such as AI Commons, trustworthy AI tools, collaborative compute infrastructure, and curated sectoral compendiums of AI use cases.
Complementing the policy and governance discussions, a major highlight of the summit will be the progress made by 12 Indian startups selected under the IndiaAI Mission to develop indigenous foundation models trained on Indian datasets and languages.
These firms are building large language models and multimodal AI systems aimed at addressing India's linguistic diversity, sector-specific needs and governance requirements.
Among them, Sarvam AI is advancing the development of large language models built with a focus on Indian languages and local requirements. The company's work spans multilingual reasoning, speech-driven technologies, and enterprise productivity tools. Under the IndiaAI Mission, Sarvam AI is contributing to the creation of a sovereign foundation model designed to handle functions such as text generation, translation, and conversational AI across multiple Indian languages.
Similarly aligning with India's multilingual priorities, Soket AI Labs is engaged in building open-source, large-scale AI systems aligned with India's linguistic diversity. Its efforts centre on multilingual and multimodal foundation models aimed at sectors including defence, healthcare, and education, with compatibility across more than 22 Indian languages.
Extending the focus from text and multimodal systems to voice technologies, Gnani AI is designing a voice-centric foundation model that prioritises multilingual speech understanding and instantaneous voice interaction. The technology is intended to support low-latency, speech-to-speech communication and is targeted at applications in customer service, learning environments, accessibility, and other public-oriented services.
In the same voice-driven domain, Gan AI is working on a multilingual AI model emphasising high-quality text-to-speech functionality. The company seeks to deliver voice synthesis performance comparable to leading global solutions, enabling use cases such as audiobooks, digital assistants, media production, and localisation for Indian-language content.
Beyond language and speech, Avaatar AI is developing specialised AI avatars customised for Indian contexts. These digital avatars can be adapted for industries like agriculture, healthcare, and governance, supporting services such as farming advisory systems, patient engagement tools, and grievance redressal mechanisms.
At the consortium level, BharatGen, spearheaded by IIT Bombay, is creating multilingual and multimodal foundation models at varying scales. The programme focuses on practical deployment across domains such as agriculture, legal services, education, finance, and healthcare, relying on datasets rooted in Indian languages and socio-cultural realities.
Adding a deep-tech and scientific computing dimension, Zenteiq is developing a multimodal foundation model named BrahmAI, aimed at enhancing engineering intelligence and scientific computation. The model is designed to aid in technical modelling, optimisation challenges, and advanced research tasks, particularly within deep-tech and engineering-heavy fields.
With an emphasis on efficiency and scalability, Gen Loop is building a family of lightweight language models supporting all 22 scheduled Indian languages. Its offerings include a foundational base model, an instruction-tuned variant, and a moderation-focused model, with an emphasis on efficiency, scalability, and safety for adoption across educational platforms, social media ecosystems, and enterprise solutions.
In the healthcare innovation space, Intellihealth, operating through NeuroDX, is developing an AI-driven system for EEG signal interpretation. The initiative seeks to enable earlier diagnosis of neurological disorders and promote progress in cost-effective brain-computer interface research, with implications for clinical care and healthcare accessibility.
From healthcare diagnostics to scientific discovery, Shodh AI is integrating artificial intelligence into laboratory operations to speed up materials research. Its platform is designed to automate experimental design and analytical assessment, helping identify novel materials such as advanced alloys and next-generation battery components.
Highlighting advances in analytical intelligence, Fractal Analytics is developing what it calls "India's first large reasoning model." The system highlights structured analytical reasoning combined with deep expertise in science, technology, engineering, and mathematics, targeting applications in diagnostics, data analytics, and complex decision-making.
At the enterprise and governance interface, Tech Mahindra's Maker's Lab is working on a foundation model tailored for Indic languages and regional dialects, alongside an agentic AI framework for enterprise and government use. The programme aims to facilitate translation, automate administrative processes, and enhance digital governance systems.
Collectively, these initiatives will be showcased at the India AI Impact Summit, underscoring India's efforts to strengthen indigenous AI capabilities while aligning innovation with linguistic diversity and national development priorities. (ANI)