Insights · tech brief

India Has 3,344 Funded Deeptech Companies. Ninety-Seven Have Reached Series B.

The received wisdom is that Indian deeptech trails the West by a few years. We tested that across eleven sectors and it holds in three. The real gap is not timing. It is that a 2014-2020 cohort of Indian deeptech converts from seed to Series B at 7.2%, against 12.3% in the UK and 18.7% in the US.

Published 06 Sept 2026

Funded deeptech companies: 3,344
Seed to Series B, 2014-20 cohort: 7.2%
Same cascade, US deeptech: 18.7%
Quantum companies past Series A: 0 of 26

There is a model of Indian deeptech that almost everyone in the market carries around. It says the West is a few years ahead, so watch what gets funded in San Francisco and London, wait three years, and you will know what gets funded in Bengaluru.

We took it seriously enough to test it. Eleven sectors, pairing the year each one inflected in US funding against the year it inflected in India.

It holds in three: AI, EV cells, and climate tech. Climate is useless as a signal because the lag is zero in both directions. It is mixed in three: defence, robotics, space. And it fails outright in five.

The five it fails in are semiconductors, quantum, drones, fusion and synthetic biology. Note which ones those are. Four of the five are sectors India has explicitly prioritised with public money, and in each case Indian policy moved before US venture capital did, not after.

The India Semiconductor Mission was approved in December 2021. The US CHIPS Act was signed in August 2022. India was eight months early. The National Quantum Mission committed 6,003 crore rupees in April 2023; the global private quantum funding inflection came in 2024. The Drone Rules and the drone PLI both landed in 2021, ahead of the Western defence-drone wave.

Anyone running the lag model on those four sectors would have been watching the wrong country.

Where the lag is real, it is not about ideas

The sectors where India does follow have a common property: they are capital-light and ungated. Nothing stops an Indian AI company from starting the week the US one does.

And the mechanism is not that Indian investors read US headlines and take cues. Foreign investors are roughly 83% of Indian startup funding since 2014. When a theme moves from the US to India, it is frequently the same limited partners rotating the same thematic allocation into a cheaper geography.

That reframes the lag entirely. It is a capital-rotation lag, not an idea-diffusion lag. It collapses to zero when the sector needs nothing but engineers, and it stretches past five years, or never arrives at all, when the sector needs local infrastructure or a local customer. Five years after Commonwealth Fusion''s 1.8 billion dollar round, India has no private fusion funding to speak of. Five years after the synthetic biology peak, no Indian wave has appeared.

So the lag is not a constant you can set to three years. It is a function of capital intensity and whether anyone in India will buy the output.

What our own registry says about the sectors policy chose

This is where it gets uncomfortable, because the numbers are ours.

Across the whole deeptech registry we track 3,344 companies with recorded funding and about 14.4 billion dollars raised cumulatively. Ninety-seven of them have reached Series B or beyond.

It is tempting to divide one by the other and call the result a graduation rate. It is not one, and we are not going to print it as one. That denominator contains companies funded last year that have had no time to graduate, companies whose only money was a grant or an angel cheque, and companies that died years ago and will sit in the data at their terminal stage forever. A ratio built on it describes the shape of our database, not the behaviour of Indian deeptech.

Computed properly it looks like this. Restrict to companies incorporated between 2014 and 2020, so every one has had at least five years to move. Count only those that reached a recorded seed round or later, so the denominator is companies that actually entered the funnel. On that basis Indian deeptech converts at 23.4% from seed to Series A and 7.2% from seed to Series B, across 525 companies.

That is the number worth arguing with. The same cascade for UK deeptech runs 29.7% to Series A and 12.3% to Series B; for US deeptech, 33.9% and 18.7%. India reaches Series B at roughly two-fifths of the US rate and a little under three-fifths of the UK rate.

Now split the capital by sector. The four frontier sectors the policy stack has been aimed at since 2021, semiconductors, quantum, space and defence, account for 237 funded companies between them, 7.1% of the funded pool. Health and life sciences, mobility, and energy account for 1,693 companies and about 7.1 billion dollars, or 49% of all capital raised.

The frontier sectors do get bigger cheques. Median raised is 8.55 million dollars in quantum, 5.42 million in space, 4.75 million in semiconductors, against roughly 1 to 1.7 million in the large sectors. Policy money, and the private capital that follows it, produces a small number of well-funded companies rather than a broad base.

It does not reliably produce graduates. Quantum has 26 funded companies and none at Series B or beyond. Semiconductors has 70 and two. Defence has 40 and one. Space is the exception, with eight of 101, and is worth studying for exactly that reason.

A correction we owe on our own quantum number

Our sector rollup reports about 887 million dollars raised across quantum. We would not use that figure without opening it, and when we did, it did not survive.

The single largest line in it is 723.3 million dollars attributed to the IITM-CDOT Samgnya Technologies Foundation, India''s official quantum communication hub under the National Quantum Mission. That is government mission funding for a national hub, sitting in a column headed private capital. It is 81.5% of the sector total.

Strip it out and Indian quantum has raised under 165 million dollars across 25 companies. Strip out the double-counting we also found, where QpiAI and QNu Labs each appear under two corporate entities carrying overlapping funding, and it is lower again.

We are reporting this rather than quietly fixing it because the correction is the finding. India''s most-publicised deeptech sector, with a 6,003 crore rupee national mission behind it, has attracted somewhere around a tenth of what the headline suggests in actual private capital, and has not moved a single company past Series A.

It also tells you what to distrust in everyone else''s numbers. IVCA put Indian deeptech funding at 2.96 billion dollars for 2025. Inc42 put the same year at 500 million. A six-fold spread between two credible sources is not a rounding difference; it is a definitional one, and it means most published Indian deeptech figures are unfalsifiable as stated.

The defence anomaly: all the IP, none of the capital

Defence is the sharpest disconnect in the dataset.

Indian defence and security patent filings went from 1,614 in 2022 to 3,421 in 2025, more than doubling in three years, and they are overwhelmingly domestic: 3,037 domestic filings against 384 foreign in 2025. On the IP side this is one of the healthiest sectors in the country.

On the capital side it has 40 funded companies and about 135 million dollars raised in total, with a median cheque of 1 million. The largest, Tonbo Imaging, has raised 55 million; Big Bang Boom Solutions 35.1 million; Sanlayan 22.4 million. After that the list falls off a cliff.

For scale, US defence tech startups raised 9.6 billion dollars in 2025 alone and 14.6 billion in the first five months of 2026.

So India is generating defence IP at volume and funding almost none of it as companies. The filings are coming from DRDO, public sector undertakings, academic institutions and individual inventors, not from a venture-backed startup base. That is a translation problem, and it is the single most addressable gap we can see in the registry.

It also points at the mechanism. iDEX''s entire five-year outlay is about 60 million dollars. US SBIR and STTR awarded 8.06 billion dollars in FY2025, 89% of it defence, and the agencies writing those cheques are also the agencies that place production orders afterwards. Every Western sector that inflected, space, defence, semiconductors, fission, had a government purchase order behind it, not investor enthusiasm. Europe is the control experiment: the EIC has funded around 706 startups and crowded in roughly three private euros per public euro, and across 2015 to 2025 exactly two VC-backed European deeptech companies remain publicly listed.

Semiconductors is the one place the IP is moving the wrong way

In quantum, space, defence and AI, Indian filings now outnumber foreign filings in our registry, in most cases by a wide margin. Space runs 191 domestic to 36 foreign. Quantum runs 109 to 26. AI runs 7,287 to 2,620.

Semiconductors is the exception, and it is not close. In 2025 India recorded 730 foreign semiconductor filings against 320 domestic, roughly 70% foreign. The foreign number is not static either: 473 in 2022, 601 in 2023, 680 in 2024, 730 in 2025.

Set that against 1.60 lakh crore rupees of approved semiconductor projects and 70 funded semiconductor startups holding 356 million dollars between them.

India is building fabs on an IP landscape that non-Indian filers are actively shaping, faster each year, while the domestic startup base in that sector remains the smallest-capitalised of the four frontier sectors. We have seen this pattern before in our battery work, where LG Energy Solution is the largest single filer in India''s energy sector. Patents get filed years ahead of factories, and whoever holds them sets the terms for whoever builds later.

What we would take from this

Stop modelling India as the West on a delay. The lag holds in three of eleven sectors, and fails in four of the five sectors India has actually prioritised. Where it holds, it is foreign LPs rotating allocation, not Indian investors taking cues. A single three-year lag parameter will be wrong more often than right.
The binding constraint is Series B, not Series A. A 2014-2020 Indian deeptech cohort converts seed to Series B at 7.2%, against 12.3% for UK deeptech and 18.7% for US deeptech. The West has the same valley; the UK has spent a decade and several billion pounds trying to cross it and can show you what did and did not work. Note the shape of the UK number, though: 12.3% is exactly what UK non-deeptech achieves too, and the UK deeptech penalty only appears at Series D, where 1.5% of the original seed cohort arrives against 2.4% for the rest of UK tech and 4.9% for US deeptech.
Announced is not disbursed, and the gap is the real leading indicator. The 1 lakh crore rupee RDI scheme had disbursed 500 crore by March 2026, half a percent of its headline, twelve months in. Anyone forecasting Indian deeptech off scheme announcements is out by one to two orders of magnitude.
Defence is the clearest opportunity and the clearest failure. Filings more than doubled to 3,421 in 2025, 89% of them domestic, against 40 funded companies. The IP exists. The companies do not. Nothing in the Western data suggests venture capital fixes this on its own; procurement does.
Semiconductors needs an IP strategy, not just a capex strategy. Seventy percent of filings are foreign and the foreign share is growing in absolute terms while fabs are still being built. The cost of entering that landscape later rises every year.

There are two ways to read all of this and both fit.

The optimistic one: India moved first on semiconductors, quantum and drones, is generating domestic IP at volume in four of five frontier sectors, and has a working IPO market the UK would trade almost anything for. The base is being laid in the right order, and capital follows evidence with a delay.

The pessimistic one: 97 companies past Series B, a national quantum sector whose headline number is mostly a government hub, and a defence sector filing 3,000 patents a year that has funded 40 companies. Public money has bought filings and pilots, not scale-ups, and the West''s own record suggests grants alone never do.

We would hold both rather than pick. But we would stop looking to San Francisco for the timing and start looking at whether anyone in India is placing a purchase order.

Methodology

Indian figures come from Deeptech Navigator''s registry as of 5 September 2026, covering 28,585 deep-tech companies and 481,614 patent applications. Graduation rates are cohort rates, not stock ratios: companies incorporated 2014-2020 by MCA record with a recorded stage of seed or later, measured to September 2026. That is 525 companies. Dividing everything currently at Series B by everything in the database would give a much lower number and would describe our database rather than the market. Patent trends run 2022 to 2025; 2026 is a partial year and is excluded.

UK and US comparators are the Royal Academy of Engineering and Dealroom, State of UK Deep Tech 2025, using a 2010-2020 seed cohort with an open window. Their Series D figures, 1.5% for UK deep tech and 4.9% for US deep tech, are the share of the original seed cohort that ever reached Series D, not a Series C to D conversion rate. Their window is longer than ours, so the direction of the gap is reliable and its exact size is not.

Two limits in our own funding data, both stated in the body rather than buried here. Sector totals include some non-venture money booked as company funding, the National Quantum Mission hub being the largest instance. And 139 brand groups appear under more than one corporate entity, covering 290 of 2,007 funded records and roughly 1.3 billion dollars of over-count. Read sector totals as upper bounds until both are fixed.

In our data

Startups

IITM CDOT Samgnya Technologies Foundation QpiAI QNu Labs Tonbo Imaging Big Bang Boom Solutions Sanlayan Krutrim

Sectors

Semiconductors Quantum Space Defence & Security Computing & AI

Sources

This briefing is AI-generated from Deeptech Navigator's patent and startup data and lightly reviewed before publishing. Treat it as a starting point, not professional advice - figures are directional, so verify before relying on any number. The platform takes no responsibility for decisions made on it.

Related briefings

tech brief

India's Semiconductor Import Bill: $9.4B From China, and Who's Actually Building the Alternative

India imported $9.4B in semiconductor devices from China in FY2024-25, 35% of its world imports in the category. Our patent and startup data shows exactly where domestic capability exists.

tech brief

ESD Protection Circuits: India's Quiet Surge in Chip-Level Resilience

As chip geometries shrink and interfaces speed up, Indian innovators are rethinking electrostatic discharge protection—balancing signal integrity, area, and adaptive control.

tech brief

India’s Advanced Packaging Innovation: Dense Interconnects, Compact Designs

From flip-chip reliability to high-bandwidth die-to-die links, Indian inventors are tackling the physical limits of chip packaging.

tech brief

Tunnel FETs in India: Engineering the Post-MOSFET Low-Power Switch

As silicon MOSFETs hit scaling walls, Indian research is reimagining the transistor with tunnel FETs that tackle power, temperature, and fabrication head‑on.

tech brief

India’s SRAM Cell Design Frontier: Power, Stability, and Speed

From low-voltage stability to leakage control, Indian inventors are rethinking memory cell topologies for a chiplet-driven world.

tech brief

India's Quantum Teleportation Frontier: Noise, Loss, and the Leap to Objects

Indian innovators tackle thermal noise, optical loss, and explore teleportation of physical objects using machine learning and plasma control, as the nation races to lead quantum communication.

Get in touch

Have a question on this - or want it researched for you?

Send a note: feedback on this briefing, a data question, or a scoped custom study on your specific market, geography or patent question. No account or card needed - we reply by email, usually within 1 business day.