Krutrim AI Brings DeepSeek R1 To Indian Developers

Ola Electric owner Bhavish Aggarwal said that Krutrim has deployed Chinese AI tool DeepSeek’s latest foundation model R1 671B on Nvidia’s H100 graphics processing units in India.

Taking to social media platform X, Aggarwal said, “While we in India should be cautious with the DeepSeek app, we can totally make use of the open source model namesake, if securely deployed on Indian servers, to leapfrog our own AI progress.” 

He also claimed that it is the most powerful open-source model available and will be offered at a rate of INR 1 per Mn tokens to Indian developers for the month of February.

This comes weeks after the AI unicorn initiated to host open source AI models of DeepSeek on its cloud platform, where five models of DeepSeek are available on the platform, ranging from 8 Bn tokens to 70 Bn trained models. 

Earlier this month, he also announced an investment of INR 2,000 Cr into Krutrim, with a commitment of INR 10,000 Cr by next year. Additionally, Aggarwal brought public notice of the roll out of a frontier artificial intelligence research lab, Krutrim AI lab.

Meanwhile, the finance ministry issued a directive, prohibiting its officers from using AI tools like ChatGPT and DeepSeek for official reasons, on grounds of over data privacy and security concerns, as per reports from a week ago.

To note, countries like Taiwan, Australia and Italy have also imposed similar restrictions on DeepSeek, citing privacy and security vulnerabilities.

This also comes at a time when the curiosity around the Chinese AI model and its potential has led DeepSeek chatbot to become the most downloaded mobile app across 140 markets, with India contributing the largest pie of new users.

Furthermore, Mumbai-based Yotta Data Services launched a B2C GenAI chatbot based on DeepSeek to offer easy access to AI solutions with focus on data security, a week ago.