MeitY to back four startups under Rs 10,000 Cr IndiaAI Mission: Report

Sarvam AI, which had earlier proposed the development of a 70-billion parameter multimodal model supporting both Indian languages and English, has already commenced work on the project. Instead of direct monetary grants, Sarvam AI is likely to receive GPU compute credits worth Rs 200 crore.

Apr 26, 2025 - 06:46

MeitY to back four startups under Rs 10,000 Cr IndiaAI Mission: Report

The Ministry of Electronics and Information Technology (MeitY) is expected to back four startups—Sarvam AI, Soket AI Labs, Gnani.ai, and Gan.ai—as the first cohort under the Rs 10,000 crore IndiaAI Mission to develop indigenous artificial intelligence models, The Economic Times reported.

Gnani.ai and Gan.ai have reportedly submitted proposals to build smaller language models.

Meanwhile. Sarvam AI, which previously proposed the development of a 70-billion parameter multimodal model capable of understanding both Indian languages and English, has already commenced work on the project. Instead of direct monetary grants, Sarvam AI is likely to receive GPU compute credits worth Rs 200 crore, according to the report.

The government is expected to announce the selected startups formally in the coming days, with Union IT minister Ashwini Vaishnaw likely to make the official declaration.

This selection comes amid growing urgency in India’s AI ecosystem, especially after China’s DeepSeek AI model garnered global attention earlier this year. The development prompted the Indian government to expedite efforts to support foundational AI models built domestically.

In support of this initiative, MeitY had announced a Rs 1,500 crore incentive fund for individuals and organisations creating AI models from scratch. The ministry received 67 applications by February 15, followed by another 120 in March.

As part of the broader infrastructure push, MeitY has also onboarded 10 GPU-as-a-service providers—including Jio, Yotta, CtrlS, Tata Communications, and NxtGen—to offer centralised compute infrastructure. These services aim to provide GPU access at rates below $1 per hour, among the lowest globally.

Edited by Megha Reddy