Ola founder Bhavish Aggarwal announced on Tuesday, February 4, an investment of Rs 2,000 crore in his artificial intelligence (AI) venture, Krutrim, with a commitment of Rs 10,000 crore by next year. He also announced the launch of Krutrim AI Lab, a frontier AI research lab, along with multiple AI models and the investment aimed at boosting India's AI ecosystem.
Also Read: Tata Communications and CoRover.ai Partner to Offer Sovereign AI Solutions in India
"While we've been working on AI for a year, today we're releasing our work to the open-source community and also publishing a bunch of technical reports. Our focus is on developing AI for India — to make AI better on Indian languages, data scarcity, cultural context, etc," Aggarwal said in a post on the X platform.
AI Infrastructure Boost with Nvidia
He also announced the deployment of India's first GB200 in partnership with Nvidia, which is expected to go live by March 2025. "We will make it the largest supercomputer in India by end of year," he said.
This announcement comes just days after Aggarwal revealed that Krutrim has deployed DeepSeek AI models on Indian services.
"India can't be left behind in AI. Krutrim has accelerated efforts to develop world class AI. As first step, our cloud now has DeepSeek models live, hosted on Indian servers. Pricing lowest in the world," Aggarwal said in a post on the X on January 31, 2025.
Also Read: India Leads in AI Adoption, Outpacing Global Average, Says BCG Report
Krutrim's Open-Source Push
The company has open-sourced its AI models and research to accelerate collaboration within India's AI community. According to the post, these resources are available on GitHub and Hugging Face.
New AI Models Announced
Aggarwal also announced the release of multiple AI models:
- Krutrim 2 LLM – The company's newly announced large language model, improving upon Krutrim 1, India's first LLM, launched in January 2024.
- Chitrarth 1 – India's first Vision Language Model for image and document understanding, built on top of Krutrim 1.
- Dhwani 1 – A Speech Language Model capable of speech translation. Ola claims this model is India's first speech-language model built on top of Krutrim 1.
- Vyakhyarth 1 – A state-of-the-art Indic Embedding model for search and retrieval-augmented generation (RAG).
- Krutrim Translate 1 – A state-of-the-art text-to-text translation model, according to Aggarwal.
Also Read: SoftBank and OpenAI Announce Partnership to Develop and Market Cristal Intelligence
Benchmark
Additionally, Aggarwal introduced BharatBench, a new benchmark designed to evaluate AI performance on Indic languages.
"In addition, since there was no global benchmark for Indic performance, we've developed 'BharatBench'," he said, sharing the technical report.
"We're nowhere close to global benchmarks yet but have made good progress in 1 year. And by open sourcing our models, we hope the entire Indian AI community collaborates to create a world-class Indian AI ecosystem. We're still learning to walk before we can run, hopefully within this year!" Aggarwal added.