The IndiaAI Mission is experiencing significant interest in its initiative to boost AI capabilities, having received 506 foundation AI proposals across three phases. Due to the overwhelming response, the submission deadline for phase 3 has been extended. The mission will announce further dates for submission of proposals and acceptance of new applications after examining the proposals already submitted.

In the first phase of approvals, Sarvam.ai has been selected to develop an indigenous foundational model with 70 billion parameters. According to cofounder Vivek Raghavan, these multi-modal, multi-scale Indian foundation models will open up a universe of applications for citizens and enterprises. The model is expected to be completed in six months.

Sarvam's model will be capable of reasoning, designed for voice, and fluent in Indian languages, making it ready for secure, population-scale deployment, according to the startup. Cofounder Pratyush Kumar mentioned that Sarvam is developing three model variants: Sarvam-Large for advanced reasoning and generation, Sarvam-Small for real-time interactive applications, and Sarvam-Edge for compact on-device tasks.

Sarvam will have access to 4,096 Nvidia H100 graphics processing units (GPU) for six months from the IndiaAI Mission's common compute cluster to train its model. They are collaborating with AI4Bharat at IIT-Madras, a leader in Indian language AI research, to build these models.