Small & smart: The Indian way in AI balances local needs, global ambition

Country's startups are carving a unique path by prioritising local, application-led innovation over the global pursuit of scale

With India’s headline-making artificial intelligence (AI) summit opening in New Delhi on Monday, attention turns to the distinctive nature of the domestic ecosystem. While global leaders like the United States (US) have set the pace, Indian startups are now showcasing a unique model of deployment and innovation. The showcase reveals a fundamental truth: India’s AI trajectory is no longer just following the global trend — it is defining its own.

 

AI firms in India last year raised $0.3 billion in 98 rounds, compared to $122.5 billion raised by US companies in 849 rounds. A total of 969 AI startups were founded in India in the past five years and 4,092 were in the US.

 

While funding for Indian AI startups remains significantly lower, investors and founders argue that the gap reflects differences in market structure rather than ambition.

 

Rajan Anandan, managing director of venture capital Peak XV Partners and startup accelerator Surge, said the US is focused on achieving artificial general intelligence (AGI) and India’s ecosystem is building an AI stack that can help the country’s 1.4 billion people. 

 

“To pursue AGI, US AI labs and Big Tech are building trillion-parameter models and committing enormous capital to compute. In recent quarters, AI and data centre investment has driven nearly all incremental US GDP growth, highlighting how capital intensive that path has become. India’s approach is different. We are seeing an explosion of applications and the rise of model companies focused on population-scale AI. Winning in India requires ultra-low pricing, local languages, and deep cultural context,” Anandan said.

 

This is evident in the problems Indian startups are working on. Company founders cite AI tools being built for Indian language customer support, voice-led interfaces for first-time internet users, precision advisory for small farmers, and AI copilots for frontline health workers. Unlike in the US, where enterprise productivity and model performance dominate the narrative, Indian startups are optimising for affordability, multilingual access, and distribution in low-bandwidth environments.

 

The AI story in India is different from that of the US, said Hemant Mohapatra, partner at Lightspeed India, another international venture capital (VC) firm.  

 

“It is less about who can build the biggest foundation models and more about who can solve the hardest real-world problems at scale. India’s current strength in AI lies in application-led innovation. We have hundreds of millions of users, unique local challenges, and a vast software services heritage that is now being repurposed into AI-driven solutions,” he said.

 

Peak XV has backed MarqVision, Mem0, Supabase, Hyperbound and other Indian companies, while Lightspeed has invested in firms, such as Sarvam AI, Composio and Emergent.

 

VC firms agree that structural differences in market depth, capital formation, and pricing power explain why US startups grow faster. The US, a $30-trillion economy, benefits from higher enterprise tech spending, faster adoption cycles, and greater willingness to pay for talent.

 

A case in point is Dashverse, an AI entertainment startup based in Bengaluru and San Francisco. The company said that while the US leads in foundational research, India is emerging as a global capital for AI’s application layer, driven by a massive, mobile-first consumer base and creative ingenuity.

 

“We have partnered with multiple production studios across India to create AI micro-dramas. This ecosystem proves that India is not just a large consumer market but has the potential to be a premier global hub for high-quality, AI-based storytelling. This strategy has allowed us to achieve the highest monetisation for AI content worldwide, with our Dashreels mobile app reaching $30 million in annual recurring revenue. For us, the future of AI is not just in the models, but in scalable experiences that deliver immediate value to the world,” said Sanidhya Narain, chief executive officer of the firm.

 

What also differentiates the Indian AI ecosystem is the ability to find willing enterprises or a mature market to accept their products.

 

“Due to faster adoption and higher willingness to pay, US companies scale faster. The US also has distribution advantages because dense clusters of early adopters are present there. Very importantly, the risk appetite for capital allocation in the US has always been much higher than in other countries,” said Bhaskar Majumdar, managing partner at Unicorn India Ventures (UIV).

 

Majumdar added that India is likely to emerge as the “application shop for the world,” as most investments are flowing into that layer. UIV has investments in companies such as TakeMe2Space, Netrasemi, and Deep Algorithm.

 

Vipul Patel, a partner at IIMA Ventures, noted that 86 per cent of the $1.8 billion raised by Indian startups between 2020 and 2025 was funnelled into AI applications. Pointing to Silicon Valley’s Perplexity AI — founded by an Indian-origin entrepreneur — he noted that the firm secured $15 million in seed funding just six months after its launch. “In the Indian context, the lack of investor interest in funding research and development for AI ventures still persists,” Patel said.

 

Ecosystem players agree that while India can capitalise on the application layer, it must address structural bottlenecks such as compute power and talent.

 

“The biggest gap is compute. India needs roughly one million GPUs over the next two to three years. Today, about 38,000 GPUs have been funded under the IndiaAI Mission, and only 15,000 are live. That means we need a 50-times increase in compute, and we need to move fast,” Anandan said, referring to processors used for high-speed parallel computing.

 

“We need to do more across compute, capital, and talent. Early signs are promising with the IndiaAI Mission set up by the government, along with similar focus from academia and industry. We must double down on our data and sovereign capability moats in AI and turn that into a long-standing advantage as a nation,” said Arjun Rao, general partner at Speciale Invest.