Our Take
Capital concentration hit a record in 2025, but the smaller-round pool didn't shrink—it just grew slower, making the real question not whether founders below $100M are dying, but whether they're being starved of velocity.
Why it matters
If you're a founder outside the mega-round cohort, you need to know whether the VC world is a zero-sum game right now (it isn't quite, yet) or whether mega-exits in AI expand the total addressable market for smaller bets. The answer shapes your fundraising timeline and your defensibility thesis.
Do this week
Founder: map your company's TAM against the white space around Anthropic and OpenAI's focus areas before your next pitch, so you can articulate why their scale paradoxically creates room for you.
Capital funneled into mega-rounds at record speed
In 2025, 70% of U.S. venture capital—more than $200 billion—landed in just 389 companies that each raised rounds of $100 million or more (per Crunchbase data). Of that, $90 billion concentrated into six companies that each raised over $5 billion. Meanwhile, around 6,000 companies split the remaining 30% ($88 billion) in smaller rounds.
The 2025 figures represent the most capital concentration on record. In 2021, the prior peak year for concentration, 60% of capital went to the $100M+ cohort. The structural shift: in 2021, more of that concentration landed in companies raising $100M to $500M. In 2025, capital flowed almost entirely to companies raising $500M and above—just 50 companies commanded that share.
Early 2026 is tracking even steeper. Through April, U.S. venture capital totals already match all of 2025, with 80% of that invested in rounds of $500M and above across just 29 companies (per Crunchbase).
The paradox: smaller rounds grew too, but slower
The harder question than concentration itself is whether the mega-round surge cannibalizes smaller startups. The data suggests no—not yet. Funding in sub-$100M rounds did not decline in 2025; it increased by roughly $8 billion to roughly the same number of companies. Rounds between $1M and $10M took a hit—down less than 10% year over year—but the overall ecosystem expanded even as capital clustered upward.
The open question is whether this dynamic holds. OpenAI, Anthropic, and similar companies must pursue extraordinarily large markets to justify their capital. That leaves surface area for smaller founders to operate in adjacent or vertical-specific use cases. But if venture capital growth decelerates while mega-rounds continue to absorb 80% of new capital, the squeeze on seed and Series A will become real.
Venture partners are betting on the expansion thesis. "There's just so much white space around that, where really interesting founders and startups can play," said Daniel Docter, managing director at Dell Technologies Capital. Madison Faulkner at NEA Ventures echoed the sentiment: "This is a moment where I'm extremely excited about betting on seed and Series A, especially in spaces that do compete with Anthropic and OpenAI." The caveat: defensibility matters. Nnamdi Okike at 645 Ventures looks for startups deeply embedded in customer workflows with proprietary data moats—not generic AI applications.
Lock down your defensibility narrative early
If you're raising below $100M, your pitch now must answer why mega-cap AI labs' resource constraints create an opening for you, not why you're a consolation prize. Faulkner's comment about Anthropic and OpenAI struggling to focus is the permission structure; use it. Show that you own a specific workflow, vertical, or data corpus they can't easily reach at scale.
For investors: the concentration trend favors repeat LP anchors in mega-rounds and managers of concentrated portfolios. Diversified seed-stage LPs face margin pressure unless they can show that their bets compound into later-stage winners. The bar for follow-on funding has risen for the middle cohort.