Something big expected as Nvidia, Sarvam team up

Nvidia has launched the Nemotron Coalition, bringing together top AI players to build the next generation of open, frontier AI models.

What’s happening: this coalition includes names like Sarvam AI (India), Mistral AI, Perplexity, LangChain, and Mira Murati’s Thinking Machines Lab. 

The goal is to pool data, compute, and expertise to build powerful AI models that are open and customisable, not locked inside closed systems.

The first outcome is a base AI model built by Nvidia and Mistral, trained on Nvidia’s DGX Cloud, which companies can tweak for specific industries, regions, or use-cases.

The why: right now, the AI race is dominated by a few closed models. Nvidia is pushing a different approach, open models that more developers, startups, and countries can build on.

Each partner brings something unique like Sarvam AI will focus on India’s languages and voice-first AI.

Big picture: the next phase of AI is shifting, and this is where India is starting to play a bigger role. 

There’s a growing push toward ‘sovereign AI’, models like Bhashini & Krutrim, that are trained on local languages, data, and cultural context.

In India, local language support is critical to reach last-mile users, especially in rural areas where English is not widely used.