The seed cheque gets bigger: Peak XV's $5 million signal
On September 29, 2026, Peak XV Partners — one of the largest venture capital firms investing across India and Southeast Asia, with more than $10 billion in assets under management — unveiled the 12th cohort of Surge, its seed-stage investing platform. The headline numbers are striking: 18 startups, more than $50 million invested by Peak XV across the group, and more than $90 million raised collectively by the companies in seed funding. But the number that matters most to founders is a quieter one: Surge 12 is the first cohort to operate under a raised per-startup investment ceiling of up to $5 million, up from the previous $3 million cap. In the world of seed investing, where the size of the first institutional cheque shapes everything that follows, a 67 percent increase in the ceiling is a statement about what it now costs to build a company.
The increase did not come from optimism alone. It came, in the firm's telling, from necessity. "The bar to raise a Series A has gone up pretty significantly," Rajan Anandan, managing director at Peak XV, said in an interview with TechCrunch. He added that the firm is also seeing more capital-intensive companies — particularly in deep tech — that need larger rounds at the seed stage simply to build their underlying technology. When the next round demands more traction, the current round has to buy more runway. The $5 million ceiling is Peak XV's answer to a market where seed-stage companies are being asked to look like Series A companies used to.
Inside Surge 12: AI, robots, and satellites
The cohort skews decisively toward artificial intelligence and capital-intensive deep tech. Among the named companies: Reinforce Labs, working on enterprise AI safety; Puralink, building robots that inspect underground pipes; and ULOOK, developing autonomous satellite systems for radio-frequency sensing. The spread — from AI safety software to subterranean robotics to orbital sensing — illustrates what "deep tech" means in 2026: it is no longer a single sector but a style of company, one where the technical risk is high, the capital requirement is large, and the payoff, if the technology works, can be enormous.
Three companies in the cohort remain in stealth: an education startup, an applied AI lab, and a medical products company. The stealth mode is itself informative — in competitive AI subfields, founders increasingly keep their exact thesis quiet until the product is ready, a marked shift from the "build in public" ethos that dominated startup culture a few years ago. The cohort also includes startups building autonomous clinical intelligence for primary care, dating products, insurance brokerage tools, AI evaluation and red-teaming software, and post-training infrastructure for open-weight AI models, according to reporting on the announcement.
The founder profiles read like a cross-section of where technical talent is flowing. Surge 12 includes a former Google director who led AI and machine-learning teams, Airbnb's first female engineer, founding engineers from Sarvam AI — the Indian AI lab working on sovereign large language models — and a former professional baseball player. The mix of big-tech veterans, proven startup operators and genuine outsiders is deliberate: Peak XV's thesis has always been that exceptional founders can come from anywhere, and the program's support infrastructure is designed to professionalize first-time founders quickly.
Surge goes global
Perhaps the most significant structural detail of Surge 12 is geographic. More than half of the 18 companies are based in India, but only five are specifically targeting the Indian market. The remaining 13 are building for customers across global markets, with founders spread across locations ranging from San Francisco to Sydney. As Anandan told TechCrunch, Surge has become more global with each cohort — and the gap between where companies are built and where they expect to find customers is now the defining feature of the program.
This is a meaningful evolution for a firm whose roots are in Sequoia Capital India and Southeast Asia. The original Surge thesis was about backing founders in the region; the current thesis is about backing founders from the region — or backed through the region's infrastructure — for the world. It reflects a broader shift in how global venture capital thinks about geography: talent is everywhere, customers are everywhere, and the accelerator's job is to connect the two regardless of where either sits. For founders in Sydney, Singapore or San Francisco, a seed program anchored in Asia with a $5 million ceiling and a track record of scaling companies globally is a credible alternative to the Silicon Valley accelerators that once had a near-monopoly on top technical founders.
The track record justifying bigger bets
Peak XV is explicit that the larger cheques are justified by results. Since its launch in 2019, Surge has backed more than 180 startups founded by entrepreneurs representing more than 18 nationalities. The firm says the 10 largest companies to emerge from those cohorts now generate more than $1 billion in combined annual revenue. That is the kind of track record that lets a firm argue — to its own investment committee and to its limited partners — that writing bigger seed cheques into more capital-intensive companies is a disciplined strategy rather than a speculative one.
The program pairs the capital with hands-on support across hiring, product and engineering, go-to-market, communications and policy — the full stack of company-building functions where early-stage founders typically have the least experience. The policy support is a notably modern addition: for deep-tech companies working in regulated spaces like healthcare, insurance or satellite communications, regulatory strategy is a core competency, not an afterthought. An accelerator that helps a satellite-sensing startup navigate spectrum licensing is offering something a generic mentorship program cannot.
What bigger seed rounds mean for the funding ladder
The Surge 12 announcement lands in a venture market that has been quietly restructuring itself. The classic funding ladder — a small seed round to find product-market fit, then a Series A to scale — assumed that the distance between those two milestones could be covered with a few million dollars and eighteen months. In 2026, for a growing share of companies, that assumption no longer holds. AI companies need compute budgets that would have been Series B-sized three years ago. Robotics and space companies need hardware, testing facilities and regulatory approvals before they can show the revenue traction that Series A investors now demand. The result is a "seed" round that looks, in size and expectations, like what a Series A used to be.
This has consequences that founders should understand clearly. A larger seed round means a higher valuation at entry, which means the Series A has to be raised at a still-higher valuation to avoid a down round — and down rounds, while no longer the stigma they once were, still complicate hiring, morale and future fundraising. The bar Anandan describes — stronger product traction, more advanced technology, greater capital efficiency before the next institutional round — does not get easier to clear just because the seed cheque got bigger. If anything, the bigger cheque raises the bar further, because investors at the next stage will expect to see what the money bought.
There is also a selection effect worth noting. When the seed ceiling rises to $5 million, the program naturally attracts more capital-intensive, technically ambitious companies — the Reinforce Labs and ULOOKs of the cohort — and relatively fewer lightweight software businesses. That is a deliberate choice by Peak XV, and it reflects where the firm believes the best risk-adjusted returns are. But it also means the seed market is bifurcating: capital-intensive deep tech can now raise $5 million at seed from a top-tier program, while a standard software-as-a-service startup may find the seed bar unchanged or even harder, as investor attention and partner time flow toward the bigger, more exciting bets.
The honest caveats
None of this guarantees outcomes for Surge 12's founders. Accelerator statistics are survivorship-biased by construction: the "top 10 alumni generating over $1 billion" tells you nothing about the median outcome across 180-plus companies, and the program's own materials do not disclose it. A $5 million seed round is a privilege and a burden — it buys runway, but it also buys expectations, and the history of venture capital is full of well-funded seed companies that could not convert capital into traction. The tougher Series A environment that justified the bigger cheques will still be there in eighteen months, and it will not grade on a curve.
For the broader startup ecosystem, though, Surge 12 is a useful data point in the ongoing debate about the future of seed investing. The direction of travel is clear: seed rounds are getting larger, more professionalized and more global, while the Series A bar keeps rising to meet them. Founders planning a fundraise in late 2026 or 2027 should calibrate accordingly — the company that could have raised a $2 million seed on a prototype in 2021 may now need demonstrable traction and a credible path to $5 million in annual revenue to raise the same round. Peak XV has not created this market. It has simply priced it.
Sources
