Nasscom Reports Indian GenAI Firms Secured $700 Million in Funding Over 3 Years

In the midst of what is termed the ‘global funding winter,’ generative artificial intelligence (GenAI) stands out as an area where investors exhibit optimism. According to estimates from the technology industry apex body Nasscom, Indian GenAI startups have garnered a total investment of $700 million over the past three years.

With over 100 genAI startups in the country, the funding influx into emerging ventures like Hindi LLM maker Sarvam AI and Adobe’s acquisition of Bengaluru-based Rephrase.ai last year serves as strong validation for the thriving AI ecosystem in India. This perspective was shared by Sangeeta Gupta, Senior Vice President, and Chief Strategy Officer at Nasscom.

Diverse GenAI offerings received funding across various domains, with text content creation (18%), chatbots and virtual assistants (18%), along with image and video generation (16%), comprising significant shares of the total investment.

Nasscom estimates reveal that AI startups, in general, garnered private investments amounting to $8 billion from 2013 to 2022. Notably, $3.24 billion attracted in 2022 alone, distributed across 1900 AI startups, marking most significant funding period the AI sector.

Also Read : MobiKwik Plans to Generate Rs 880 Crore Through Fresh IPO Offering

The past year saw significant breakthroughs in GenAI innovation, as both tech majors and startups developed large language models trained on Indic language

Conversational AI startup Corover.ai unveiled BharatGPT, 7 billion parameter model trained in 14 Indian languages across text, voice, and video. Simultaneously, Sarvam AI introduced the inaugural open-source Hindi language model, named OpenHathi-Hi-0.1, constructed on Meta’s LlaMa 2-7B model.

Mobility unicorn Ola unveiled the Krutrim LLM, capable of comprehending 22 Indian languages and generating responses in 10 languages. Meanwhile, IT services leader Tech Mahindra actively engaged ‘Project Indus,’ developing an LLM trained in Hindi and 37 Indic dialects.

Additionally, companies such as Flipkart and SaaS provider Ozonetel are delving into small language models trained on more compact datasets, customized for specific domain use cases.