Sarvam.ai, one of the key beneficiaries of the IndiaAI initiative, is training large-scale language models using a compute infrastructure comprising 4,096 Nvidia H100 GPUs over six months. The startup was selected in the mission’s first phase, receiving a substantial portion of the allocated compute subsidy to support its development of indigenous AI capabilities tailored to India’s diverse linguistic and cultural landscape.

The company has now confirmed that these models, once trained, will be released under open-source licenses, enabling developers, researchers, and enterprises to access and build upon them freely. The move is aimed at accelerating the creation of domain-specific AI tools, improving language inclusion, and encouraging ecosystem-wide collaboration across public and private sectors.

The open-source commitment includes weekly model drops, detailed documentation, and public APIs. Sarvam.ai plans to publish comprehensive updates, including performance benchmarks for Indian languages, reasoning tasks, and other capabilities, inviting feedback and contributions from the wider AI community.

This initiative aligns with the broader objectives of the IndiaAI Mission, which was launched with a budgetary allocation of ₹10,000 crore to support the development of foundational AI technologies within the country. The mission focuses on building sovereign AI capabilities by funding access to high-end GPUs, incentivising model development, and promoting open innovation.

India’s push for open-source AI models is being driven by the need for language equity, cultural contextualization, and technology democratisation. By opening up access to foundational models, the government aims to empower Indian startups, academia, and developers to create AI tools that are affordable, locally relevant, and scalable.

Sarvam.ai’s leadership emphasised that the financial support received under the mission is structured as an equity-based investment rather than a grant. The startup aims to balance commercial sustainability with public interest by delivering models that can support critical use cases in education, governance, agriculture, healthcare, and more.

The announcement is also seen as a response to growing public interest in ensuring that government-supported AI infrastructure remains accessible and transparent. With over 500 proposals submitted under the IndiaAI Mission—more than 40 of which focus on large language models—the move toward open-source AI represents a strategic shift in how India envisions the future of artificial intelligence.

Looking ahead, Sarvam.ai intends to contribute to the creation of AI systems that are multilingual, inclusive, and aligned with Indian values and priorities. The decision to open-source these models sets a precedent for other mission-aligned startups and reinforces India’s positioning as a global leader in responsible and equitable AI development.

Be a part of Elets Collaborative Initiatives. Join Us for Upcoming Events and explore business opportunities. Like us on Facebook , connect with us on LinkedIn and follow us on Twitter, Instagram.

"Exciting news! Elets technomedia is now on WhatsApp Channels Subscribe today by clicking the link and stay updated with the latest insights!" Click here!