The Centre has announced a new cohort of eight entities chosen to spearhead the development of foundational large language models (LLMs) under the ₹10,372 Cr IndiaAI Mission. This landmark expansion aims to catapult India into the league of global AI model creators by fostering powerful, India-centric AI innovations built on local datasets and languages.
Meet the New Cohort: Company-by-Company Highlights
- Tech Mahindra: A major Indian IT firm, Tech Mahindra brings its global reach and robust AI division to the table. It’s been investing deeply in generative AI, having launched its own in-house LLMs and enterprise-grade AI platforms. In IndiaAI, it is expected to leverage its expertise in telecommunications, manufacturing, and banking automation to build tailored, industry-grade language models.
- Fractal Analytics: A Mumbai-based AI powerhouse, Fractal recently launched a 14B parameter open-source LLM (“Fathom-R1-14B”), claiming industry-leading math reasoning—surpassing some of OpenAI’s smaller models. Fractal’s IndiaAI work targets “reasoning models” for STEM, coding, and advanced business analytics, with open-source, cost-efficient approaches.
- Avataar AI: This VC-backed startup from Bengaluru specializes in visual AI, 3D computer vision, and B2B LLMs. Avataar is expected to focus on sector-specific foundational models, leveraging its strengths in retail, logistics, and smart infrastructure deployments for the IndiaAI Mission.
- Zeinteiq Aitech Innovations: Based out of Bengaluru’s Indian Institute of Science ecosystem, Zeinteiq (or ZenteiQ.ai) pioneers “Science Foundation Models”—large-scale AI for scientific simulations, engineering, and strategic sectors like defense and aerospace. Their focus is on “physics-informed” models, scientific reasoning, and secure, sovereign AI for national gains.
- Genloop Intelligence: With a presence in New Delhi and San Francisco, Genloop is building a multilingual compliance LLM, explicitly trained on Indian culture, law, history, Parliament records, and regional media. Their 22-language support positions Genloop for high-impact, India-focused regulatory and governance use cases.
- NeuroDX (Intellihealth): This health-AI specialist blends proprietary neural architectures with medical expertise. Its AI platforms are used for diagnostics and analytics, and under IndiaAI, NeuroDX is set to contribute biomedical LLMs for clinical decision support, public health, and personalized medicine.
- Shodh AI: Shodh is a new startup selected to develop foundational models for India, emphasizing open research and academic-industry collaboration. Its charter includes domain-specific LLMs and dataset initiatives for languages, education, and healthcare.
- IIT Bombay (BharatGen Consortium): Anchored by IIT Bombay, BharatGen aims to build a trillion-parameter LLM—India’s largest effort to date—with nearly ₹990 Cr in government backing. The consortium brings together multi-institutional expertise and is the Mission’s flagship academic project.
Previous Rounds and Other IndiaAI Recipients
Earlier rounds saw the selection of Sarvam AI (which raised $53.6M and launched the hybrid “Sarvam-M” model), Gnani.ai (speech AI and Indian voice interfaces), Gan AI, and SoketAI—each focused on either sub-1B parameter models, speech, or NLU optimized for Indian dialects and enterprise use.
IndiaAI Mission: Funding, Scale, and Objectives
- Funding & Scale: Conceived in March 2024, the IndiaAI Mission provides a five-year, ₹10,372–20,000 Cr outlay, with high ambitions to democratize AI infra and boost indigenous datasets, sovereign LLMs, and foundational model R&D.
- Infrastructure: IndiaAI is deploying 38,000 GPUs nationwide for low-cost access (~₹65/hr), distributing compute via partners like Yotta and NxtGen Cloud, and building 500–600 data labs for startups, research, and public applications.
- Framework & Vision: The government will soon release an AI development and deployment framework, and is onboarding multiple industry and research partners across verticals like healthcare, agriculture, governance, and national language processing.
Read this: India Get Its First ‘Made in India’ Semiconductor Chip at Semicon India 2025
Impact and Outlook
This new cohort—mixing IT giants, R&D-driven startups, and India’s top academic talent—demonstrates IndiaAI’s multi-pronged approach: sovereign capability, deep Indian language/sectoral support, and public-private synergy. As India’s computing and model-building power grows, it eyes a global leadership stake in open, democratized, and responsible AI
