India’s venture capital conversation in 2025 and 2026 has been dominated by the big numbers — unicorn listings, AI mega-rounds, and flagship funds crossing $500 million. But the more consequential story is happening at the bottom of the funnel, where first cheques are being written into companies that most investors have never heard of, in sectors that may or may not produce returns for a decade. Pre-seed and seed funding in India is not just active — it is structurally changing. The investors writing those early cheques, the sectors attracting them, and the startups that are getting funded (and the ones that are not) tell a specific story about where India’s next generation of companies is being built, and by whom.

The macro picture — early stage is bucking the broader trend

India’s startup ecosystem raised approximately $9.1 billion in total VC funding in 2025 — up 23% year on year from the 2023 trough, according to Nasscom. In H1 2026, that number was $5.2 billion across 501 deals. But the aggregate figure obscures the most important structural shift happening underneath it: late-stage funding has tightened sharply while early-stage is accelerating.

Early-stage tech startup funding in India grew 33% in FY26 compared to FY25, even as total tech startup funding dropped 18%. In Q1 2026 alone, Indian startups raised close to $4 billion, with early-stage capital crossing $1 billion across 122 seed and 41 pre-seed rounds. Seed stage funding was up 58% year on year per Inc42. In 2025, more than 1,300 of the 2,130 startups that raised funding were at early stage — seed or Series A — collectively raising $3.85 billion.

The interpretation is clear: institutional investors are writing fewer, larger cheques at growth stage while deploying more capital, across more bets, at the earliest stages. The filter is stricter than it was in 2021, but the capital is there and actively moving. As one market analysis put it: the money is there; the filter is just stricter than it was two years ago.

Metric Figure
Total VC deployed in India — 2025 ~$9.1 Bn (up 23% YoY) · 1,000+ deals
Total VC deployed in India — H1 2026 $5.2 Bn across 501 deals (down 9% YoY but deal count up 7%)
Early-stage funding growth — FY26 +33% YoY even as total tech startup funding fell 18%
Seed stage growth — 2025 +58% YoY per Inc42
Q1 2026 early-stage deal count 122 seed rounds + 41 pre-seed rounds
Typical pre-seed cheque in India — 2026 ₹2.5 Cr – ₹14 Cr ($300K – $1.7M)
Typical seed cheque in India — 2026 ₹4 Cr – ₹15 Cr ($500K – $1.8M) · average ₹7–8 Cr
Micro-VC funds active in India — 2026 250+ (up from ~200 in 2024) per IVCA

The investors — who is writing India’s earliest cheques

The pre-seed and seed investor landscape in India has matured significantly. It now spans four distinct tiers: dedicated pre-seed micro-VCs, institutional seed funds, generalist early-stage funds with seed programmes, and solo GPs. Each tier operates with a different cheque size, thesis, and diligence model.

The dedicated pre-seed funds

100X.VC is the most systematic pre-seed investor in India by volume. Founded in 2019 by Sanjay Mehta, Ninad Karpe, Shashank Randev, Yagnesh Sanghrajka, and Vatsal Kanakiya, it was the first Indian VC to invest using iSAFE notes — India’s version of the US SAFE note. The firm deploys ₹1.25 crore per startup for 15% future equity, running cohort batches of 17–18 companies twice a year. It screens more than 10,000 applications annually. As of early 2025, it has backed 199 companies across 12 cohorts, with portfolio companies including Agnikul Cosmos (now a public company), Emo Energy, and Dezy. The model is deliberately high-volume: not every bet will work, but the iSAFE structure keeps deployment costs low and allows the portfolio to be built across sectors and geographies without the overhead of a full diligence process at the initial cheque stage.

Better Capital is India’s largest pre-seed fund by portfolio volume. A solo GP-led fund, it has backed more than 200 companies, writing cheques typically in the $100K–$350K range. It runs lean — the solo GP model means every investment decision is made by one person with full conviction, not by committee. The fund operates across all sectors with a technology-first filter.

Together Fund focuses on pre-seed and early seed, writing cheques from $200K to $500K in consumer internet, SaaS, and fintech. It requires a working MVP and early customer conversations — beyond that, it is thesis-flexible.

PointOne Capital is a pre-seed micro-VC with 70+ investments, focused on tech-led and deep-tech teams at the earliest stage. It has positioned itself as one of the first institutional cheques for founders coming out of top engineering and management institutions.

All In Capital is a ₹300 crore pre-seed fund that markets itself as wiring capital first — before other investors commit. Founded by former First Cheque partners, it focuses on speed of decision-making as a competitive advantage in a market where pre-seed founders often lose time to slow diligence processes.

The institutional seed funds

Blume Ventures is the most active institutional seed investor in India by deal count. The firm participated in 34 funding deals in 2025 alone, with the majority at seed and Series A stages. It raised $175 million in the initial close of its fifth fund, targeting $250–275 million total. Portfolio companies include Unacademy, GreyOrange, Dunzo, Yulu, and Namma Yatri. The fund’s 2025 bets include D2C fashion brand Freakins, personal care brand Plush, and deeptech imaging startup Optimized Electrotech — a characteristically diverse set that reflects Blume’s sector-agnostic, founder-first thesis. The fund tracks weekly activation data, founder velocity, and capital efficiency at the pre-investment stage.

Stellaris Venture Partners has made 61 seed-stage investments with an average round size of $2.56 million, making it one of the highest-conviction seed investors in the country — it writes meaningful cheques and actively supports portfolio companies post-investment. The fund, raised at $300 million (Fund III), is deploying $100–150 million into AI startups specifically. Portfolio includes Darwinbox, Cashfree, and LeadSquared. In 2025 it made 16 investments; as of August 2026 it has made 6. Its most recent first investment was in Pinegap.

3one4 Capital closed 40 deals across sectors in 2025. Founded by Pranav Pai and Siddarth Pai, it manages $200 million across Fund IV and has backed 80+ companies including Darwinbox, Licious, Blue Tokai Coffee, and Tracxn. The fund operates from pre-seed through Series A and leads or co-leads with disciplined follow-ons. It tracks founder velocity and business model clarity as primary underwriting criteria.

India Quotient is one of India’s most founder-friendly early-stage funds, writing cheques from $500K to $3 million. Its portfolio is skewed toward consumer internet, vernacular platforms, and B2B SaaS. Notable investments include ShareChat, Trell, and Bombay Shaving Company.

The generalist funds with pre-seed programmes

Accel India runs Atoms — its pre-seed programme — to identify companies in AI and other emerging sectors before they’re ready for institutional seed rounds. In 2026, Accel partnered with Prosus to launch Atoms X, specifically for deeptech startups, backing companies including Praan, Dognosis, and EtherealX. Accel just closed its ninth India fund at $550 million — its largest India vehicle — with an explicit AI and deeptech focus. The Atoms programme gives Accel proprietary early access to companies that competitors only see at Series A.

Kalaari Capital, led by Vani Kola, leads early-stage deal volume alongside Blume and has backed Dream11, Cure.fit, and Snapdeal at seed or pre-Series A. The fund writes $500K to $5 million cheques across consumer internet, fintech, healthcare, and SaaS.

The sector-specific early-stage funds

Speciale Invest focuses exclusively on deep-tech — space, defence, robotics, and semiconductors. Its ₹600 crore Fund III covers exactly the sectors that are attracting disproportionate government and institutional attention in 2026. Recent investments include Netrasemi (edge AI chips, ₹107 crore Series A lead) and several space and defence startups.

Inflexion Point Ventures focuses on consumer brands and D2C at early stage. Sauce VC (which backed Bruno Milano this week) focuses on consumer and D2C at seed. Chiratae Ventures has backed Peeko at Series A and has a strong consumer track record including Myntra and FirstCry. Rainmatter (Zerodha’s VC arm) focuses on health and climate tech, with consecutive investments in Solinas Integrity and Lissun in August 2026.

Which sectors are getting funded — and which are not

The sector distribution of pre-seed and seed capital in India in 2025–2026 reflects three overlapping dynamics: the AI wave, government policy tailwinds in deeptech and defence, and the structural digitisation of India’s consumer and enterprise markets.

AI and deeptech dominate the narrative and increasingly the cheques. Indian AI startups raised $676 million in H1 2026 — a 4x increase from $162 million in H1 2025. AI funding at seed is now moving in two directions simultaneously: application-layer AI (AI-native SaaS, vertical AI tools, enterprise AI agents) and infrastructure-layer AI (chips, edge inference hardware, AI data centres). Speciale Invest and Stellaris are explicitly deploying into deeptech. Accel’s Atoms X programme was designed specifically for this cohort. The government’s IndiaAI Mission, with $1.2 billion committed, has influenced 66% of institutional investors’ AI investment thesis.

Fintech remains the largest single sector by deployed capital — $1.3 billion in H1 2026 — but the early-stage action is more nuanced. At seed, the deals that are getting done are in infrastructure fintech (lending tech, credit decisioning, collections AI), wealthtech (retail investment platforms), and insurtech. Generic payment or wallet plays are no longer fundable at seed without a specific distribution advantage.

Consumer and D2C are bifurcating sharply. Well-differentiated consumer brands with genuine product moats and quick commerce distribution — BlissClub, Bruno Milano, Peeko — are getting funded. Undifferentiated consumer plays competing on marketing spend alone are not. The era of funding a D2C brand purely on the strength of its category opportunity is over; investors want to see the product differentiation before the category analysis.

Climate tech and EV infrastructure are attracting a growing share of early-stage capital, driven by government policy, large corporate commitments, and a cohort of investors — Rainmatter, Ecosystem Integrity Fund, Aeravti Ventures — whose thesis is explicitly climate-positive. Battery Smart, Yulu, and a range of smaller EV infrastructure startups have received early backing from this cohort.

Healthtech is growing but fragmented. The sector is attracting capital across sub-segments — mental health (Lissun, Mave Health), diagnostics (Dognosis), medical devices (Voxelgrids, vTitan), and health AI. But the path from seed to Series A in healthtech is longer and more expensive than in SaaS — clinical validation, regulatory navigation, and hospital partnerships all take time and capital that typical early-stage fund timelines struggle to accommodate.

Defence tech is the newest fundable category. Policy changes in 2023–2024 opened Indian defence procurement to startups, and a handful of early-stage defence tech companies have since raised from Speciale Invest, Bluehill.VC, and a new cohort of defence-focused angels. This category was essentially unfundable three years ago; in 2026 it is attracting dedicated institutional capital.

Edtech is recovering, slowly. The Byju’s collapse — the most consequential startup failure in Indian history — scarred the sector and made it functionally unfundable for two years. In 2025–2026, a more cautious edtech cohort focused on B2B skills development, upskilling for enterprise, and vernacular content platforms is beginning to attract seed capital again. The B2C tutoring model that Byju’s represented is not.

What is not getting funded — the filter has tightened materially

Understanding what is not getting seed funding in India in 2025–2026 is as instructive as knowing what is. Several patterns have emerged from the data on shutdowns and failed fundraises:

Application-layer AI without defensibility. The most common failure mode in 2025–2026 at the seed stage is the AI application that could be replicated by a well-prompted ChatGPT wrapper. Investors have become sophisticated enough to ask: what happens to your product when OpenAI or Anthropic ships the feature you’re building? If the answer is “our product becomes irrelevant,” the round does not close. Amogha AI — which we covered in an earlier Founder Graph analysis — is a clear example: the thesis was correct, the product was technically ambitious, but the proof points weren’t enough to overcome investor hesitation about long-term retention and defensibility.

Consumer plays without a distribution moat. A consumer product that reaches customers only through paid digital advertising is not fundable at seed in 2026 unless the unit economics are demonstrably excellent from day one. The era of “we’ll figure out distribution at Series A” has ended. Investors want to see a specific, defensible customer acquisition mechanism — quick commerce shelf presence, community-led growth, enterprise sales, or a distribution partnership — before writing the first cheque.

Fintech without a regulatory path. The RBI’s increasing scrutiny of consumer lending, P2P platforms, and payments has made regulatory navigation a first-order question at seed. Founders who cannot answer “how do you operate within the regulatory framework and what happens if that framework changes?” are finding it very difficult to close pre-seed and seed rounds in fintech.

B2C social and content plays. With ShareChat struggling and Koo having shut down, the investor appetite for Indian social media and content platforms has effectively evaporated. The combination of deep-pocketed competition (Instagram, YouTube), regulatory uncertainty, and poor monetisation track records makes this sector structurally difficult to fund.

Hardware without a manufacturing plan. India’s government is actively incentivising domestic manufacturing through PLI schemes and the semiconductor mission. But investors have learned from a generation of hardware startups that importing components and assembling in India is not a sustainable competitive advantage. Hardware startups that cannot articulate a credible domestic manufacturing path — like Wippi’s explicit India-first manufacturing commitment — are finding seed rounds difficult to close.

The failure picture — what shutdowns at seed reveal

India saw 11,223 startup closures in the first ten months of 2025 — a 30% increase from 8,649 closures in 2024. That is approximately 37 startups failing every single day. The profile of 2025–2026 shutdowns is different from prior years in an important and specific way: these are not the seed-stage misfires that fold quickly when the idea doesn’t find a market. Many are companies that raised meaningful capital, built real products, and found real users — and still could not close their next round.

The common failure patterns from the data:

The bridge round problem. Startups that raised pre-seed capital and built to a point of genuine traction are discovering that bridge rounds — the capital needed to get from pre-seed proof points to the metrics required for a seed round — are very difficult to close in 2026. Investors who might previously have written a bridge cheque on relationship and promise are now waiting for specific milestones before committing. Companies caught in this gap — past the idea stage but not yet at seed-ready metrics — are the most common failure mode.

Single-client dependency. Several 2026 shutdowns trace directly to a single enterprise client loss or delayed payment that wiped out runway for a company with no reserves. In a tighter funding environment where bridge rounds are harder to close, a small revenue shock can become existential quickly. Pre-seed founders are being advised — correctly — to diversify revenue before raising seed, not after.

The zombie cohort. Perhaps more consequential than the official shutdowns is the “zombie” cohort — companies that are technically alive but operating with two or three employees and no meaningful growth trajectory. Multiple analysts have noted that the real failure rate in Indian startups is materially higher than the official closure numbers suggest, once zombie companies are counted. These are founders who have not officially shut down but are not building — stuck, unable to raise, and unwilling to return capital.

subtl.ai — a Hyderabad-based GenAI startup building infrastructure for document processing and retrieval — is the most instructive case study from this period, and its founder Vishnu Ramesh’s account of what went wrong is worth examining in full.

Ramesh and his co-founder, chief scientist Prof Manish Shrivastava, started subtl.ai in 2019 — five years before RAG (Retrieval-Augmented Generation) became a mainstream term — building infrastructure for complex document processing and transformer-led retrieval. They had real customers: defence PSUs, two airports, a US insurtech, and the State Bank of India. They published a patent and a research paper on their approach to graph construction and vector embeddings that demonstrably outperformed OpenAI on 20,000 RBI banking regulations. They raised $200K from angel investors. And they still shut down.

Ramesh’s post-mortem on LinkedIn is unusually honest. He identified four failure modes, none of which were about the technology:

Lack of market focus. “We wanted to do OpenAI-like things for the RAG ecosystem, but I got stuck handling customers from wildly different domains with wildly different use cases. It was just not repeatable, so customers gave no shits about our other portfolio of work we had done.”

Lack of infrastructure GTM. The company built APIs for RAG but made almost no effort to help developers adopt them. “All we did was put a message on our website saying ‘yo reach out if you wanna use our APIs.'” No open-source integrations, no framework partnerships (LlamaIndex, Portkey), no developer documentation, no PLG motion.

Founder-role mismatch. “I’m a better CTO than a CEO. I love building in AI with all my heart. I have learnt to enjoy selling, but true sales chops come from deep domain expertise. Being a vagabond across industries helped me learn a lot about the world, but still not enough to convince people to pay us.”

Investor conversations are not commitments. “Some investors flirt a lot with founders, but it doesn’t mean shit until they give you a term sheet. I had made decisions assuming that funding budgets would come in based on conversations I had with a few good funds out of Hyderabad. This became a big reality check for me. Be ruthless with your time.”

The subtl.ai story carries a specific lesson that extends well beyond one company’s experience. A technically credible team, real customers, genuine IP, and an angel cheque are not sufficient conditions for survival at pre-seed if the go-to-market is unfocused and the fundraising process treats investor interest as equivalent to committed capital. Ramesh’s fourth point — that investor conversations without term sheets are worthless as planning inputs — is the most practically important lesson in his account, and it is one that the current environment makes more urgent, not less.

Ramesh ended his post with characteristic clarity: “I failed my team and investors more than they failed me for sure.” That kind of accountability, publicly stated, is rare and worth noting — and it explains why the startup ecosystem increasingly views failure with the right lessons learned as a credential rather than a disqualification.

The geography picture — where early-stage capital is concentrating

Bengaluru continues to capture approximately 40% of all early-stage deals, driven by its engineering talent density, the concentration of angel investors who have exited from Infosys, Flipkart, and the first generation of Indian unicorns, and the ecosystem of accelerators, incubators, and co-working communities that make early-stage company formation faster and cheaper than anywhere else in India.

Delhi NCR is the second-largest hub, raising $2.2 billion across 224 deals in 2025. Seed deal counts held flat but cheques got smaller — reflecting a more cautious investor cohort in a market where government proximity matters more than in Bengaluru. Mumbai is third, with a concentration of fintech and consumer brands benefiting from proximity to financial services infrastructure and the city’s historically strong angel network.

The more interesting geography story is the rise of tier 2 hubs. Chennai, with five unicorns and a strong SaaS and manufacturing base, is producing increasingly credible early-stage companies. Hyderabad is attracting deeptech founders, partly due to T-Hub and the state government’s active startup policy. Pune has eight unicorns and a product-first culture that is producing B2B SaaS companies with strong enterprise sales credibility.

The new fund formation wave — what it means for founders

The micro-VC ecosystem in India has expanded from approximately 200 active funds in 2024 to more than 250 in 2026, per IVCA data. This proliferation is driven by two forces: domestic limited partners who are increasingly comfortable backing smaller, thematic funds that promise more attention and better access; and a generation of experienced operators and angels who are formalising their investment activity into fund structures.

Several significant new fund formations in 2025–2026 are worth noting for early-stage founders:

Accel Fund IX ($550 million) — the largest India-focused fund Accel has ever raised. Explicit AI and deeptech focus. The Atoms and Atoms X pre-seed programmes give Accel proprietary deal flow at the earliest stage.

Elevation Capital Fund IX ($500 million) — targeting consumer tech, consumer brands, fintech, enterprise AI, frontier tech, and healthcare. One of India’s most experienced funds doubling its war chest.

Blume Ventures Fund V ($250–275 million) — India’s most active seed-stage fund expanding its corpus for its fifth vehicle.

AUM Ventures India Innovation Fund (₹750 crore, first close at ₹225 crore) — investing primarily at pre-seed and seed in deeptech and frontier technology. A new entrant with explicit early-stage focus.

Bluehill.VC (₹400 crore) — frontier tech fund, final close. Focus on defence, deep manufacturing, and frontier engineering.

Mirae Asset Venture Investments India (₹1,800 crore, first close at ₹1,125 crore) — Series B to Series D, but signals institutional capital deepening across the stack.

For founders, this fund formation wave has a specific implication: the investors writing pre-seed and seed cheques in 2026 have more capital, a longer time horizon, and clearer sector theses than at any point in Indian VC history. The competition for the best early-stage deals is real and intensifying. The opportunity for founders with genuine differentiation — product depth, sector expertise, a specific distribution insight — to close pre-seed and seed rounds has never been better. The opportunity for undifferentiated founders has never been worse.

What this means for founders raising pre-seed or seed in 2026

Several clear principles emerge from the data on who is getting funded, who is not, and what the active investors are looking for:

Domain expertise is now the baseline. Investors want founders who have lived the problem — someone who worked in logistics for years and then starts a logistics startup is more convincing than someone jumping in with no experience. This is not a new principle, but in 2026 it has become a near-universal filter at pre-seed and seed. First-time founders without sector depth are finding it materially harder to close rounds than founders with operational experience in the problem they’re solving.

The market size story has been superseded by the market entry story. Investors no longer want to hear that the market is $10 billion — they want to hear exactly who you are selling to, how you acquired your first 10 customers, and what the unit economics of that acquisition look like. A narrow, believable beachhead with demonstrated retention is more fundable than a large market with no early customers.

Capital efficiency is now explicitly tracked from day one. In 2021, investors were comfortable funding companies that would figure out capital efficiency later. In 2026, the tracking starts from the first conversation. Blume’s diligence explicitly covers capital efficiency. 3one4 tracks founder velocity and kill/scale criteria. The fastest way to lose a pre-seed meeting is to be unable to articulate what you will stop doing if a specific test does not work.

The right investor matters as much as the cheque. With 250+ micro-VCs active in India, founders are increasingly able to choose investors who have specific expertise in their sector, specific distribution relationships, and specific follow-on capital access. A cheque from Speciale Invest into a defence tech startup is worth materially more than a cheque from a generalist angel with no defence network, even at the same valuation and cheque size. The investor’s network, sector knowledge, and follow-on capacity are part of the product the investor is selling.

Sources: Inc42 Indian Tech Startup Funding Report H1 2026, Nasscom, Tracxn, YourStory, Peony Investor Database, Backrr VC Database, IVCA, Economic Times, The Captable, Entrackr, CIOL. Data as of August 24, 2026.