Indian AI startups raised $1.6 billion across 81 rounds in the first nine months of 2026, compared with $677 million a year earlier, according to Tracxn data shared with ET AI. However, the number of funding rounds fell 43% from 142, while the number of companies that raised money dropped 44% to 70 from 126.
Four large transactions – sovereign AI infrastructure provider Neysa's $600 million round, generative AI platform Sarvam's $255 million round, AI infrastructure provider Yotta's $150 million funding and AI-powered app-building platform Emergent's $130 million round – accounted for 72% of India's AI funding during the period.
"AI funding has risen in dollars but not in the number of deals and almost all of the rise comes from AI infrastructure rounds," said Neha Singh, co-founder of Tracxn.
Singh said funding would have declined without these large rounds.
Globally, investors are also putting much of their money into a few AI companies, although the amounts are larger. AI companies raised $519 billion globally up to mid-September 2026, according to the State of AI Report 2026 by venture capital firm Air Street Capital and data platform Dealroom.
The US accounted for $393 billion, or 76% of the total, followed by China with $59 billion and Europe with $40 billion. OpenAI and Anthropic together raised $217 billion, accounting for about 42% of global AI funding during the period.
In India, AI infrastructure raked in $1.2 billion across 19 rounds in the first nine months of 2026, accounting for 75% of total AI funding, according to Tracxn. Funding increased 8.3-fold from $145 million a year earlier, even as the number of rounds fell from 29.
Neysa and Yotta together raised $750 million for AI cloud and computing infrastructure, while Sarvam raised $346 million across two rounds ($255 million in June and $90 million in August) to develop foundational models.
Private equity firms, technology companies and other strategic investors are now putting money into AI infrastructure, alongside venture capital firms that have traditionally funded software startups. Blackstone led Neysa's round, while Sarvam's funding included investments from HCL Technologies and Nvidia, and Yotta's round included the Serum Institute of India.
"Building, interconnecting, cooling, and operating high-density clusters reliably requires specialised skills across power, cooling, networking and distributed systems. Expanding the engineering capability is essential as deployments scale," he told ET AI.
While infrastructure attracts the largest investments, companies building AI applications are raising less money, even as investors see opportunities in manufacturing, healthcare, energy and other industries.
Indian AI application startups raised $390 million across 62 rounds in the first nine months of 2026, down 27% from $532 million across 113 rounds a year earlier, according to Singh.
Funding for vertical AI, which serves specific industries, fell 87% to $41 million from $319 million, while consumer AI applications attracted just $1 million across three rounds. Enterprise AI applications were an exception, attracting $348 million, up 70% from a year earlier, although Emergent and DESCRIPTOR???? Freehand together accounted for 79% of that funding.
"Most of what gets pitched to us is still built on models trained elsewhere, and a good share of it is a thin layer on top," he told ET AI.
He said his firm was meeting more founders who had worked in pharmaceutical plants or chip design teams and were using AI to solve problems they understood well.
Ankit Kedia, founder and lead investor at Capital-A, also sees opportunities in manufacturing, industrial automation, robotics, energy and supply chains.
"The real value of AI will emerge when it improves productivity, reduces operating costs, enhances resource utilisation and delivers measurable business outcomes," he told ET AI.
Finding paying customers can be just as difficult as raising money, with Vallabhaneni saying investors increasingly ask who pays for a product and why they would continue after the next model release.
"Indian enterprises love to pilot. They are much slower to sign a real purchase order, and a startup can lose a year in pilots that go nowhere," he said.
Shweta Rajpal Kohli, president and CEO of industry body Startup Policy Forum, said startups still faced difficulties in accessing affordable computing, quality datasets, specialised research talent and investors willing to fund them over the long term. Getting businesses to adopt and pay for their products remained another challenge, she said.
Krishna Khandelwal, co-founder and CEO of conversational AI startup Hunar.AI, said enterprises increasingly wanted products that could carry out work within their existing systems.
"Enterprises want AI that can execute complete workflows and integrate into their existing systems," he told ET AI.
Vallabhaneni said startups could build a stronger business through data they owned, specialised engineering knowledge, regulatory approvals and products that worked closely with customers' existing systems.
Kohli said India had a large opportunity in AI applications, even though it did not yet compete with the US or China in the most advanced foundational AI research.
"Some of the most valuable AI businesses will be those that understand a particular industry, solve a problem better than anyone else and build products that customers are willing to pay for," she said.
