Series B gets bigger and rarer: $48B, fewer deals, and the trap it sets for seed

On 29 October 2026, Crunchbase News published a simple but uncomfortable measurement: US Series B-stage funding reached roughly $48 billion in 2026, the highest in years, while round counts stayed below year-earlier levels. Same stage, more money, fewer open doors.

More capital, narrower gate

The structure behind the $48 billion matters more than the headline. Per Crunchbase, at least 110 companies raised $100 million or more at Series B this year, about a quarter of deals in the sample. At the other end, rounds below $30 million - roughly half the market in 2021-25 - now account for under a third. The largest transaction was the $12 billion raise by Prometheus, the AI company co-founded by Jeff Bezos; Crunchbase notes the company did not itself label it a Series B, though other reporting did. Nuclear energy startup Valar Atomics secured $1 billion in Series B capital across tranches, led by Sequoia Capital.

About two-thirds of this year's Series B dollars went to AI-focused companies, per Crunchbase. The concentration is not stage-specific: global venture funding hit a record $510 billion in H1 2026, more than all of 2025's $440 billion, yet OpenAI and Anthropic alone absorbed over 40% of it. The market runs at two speeds - a mega-tier on near-unlimited capital, and everyone else on an ordinary budget.

• Series B is no longer one rung on a ladder; it has split into two different products - nine-figure rounds for leaders, increasingly scarce small rounds for everyone else.
• Later-stage investors are paying for certainty, not potential, concentrating capital in a handful of already-legible companies.
• Companies outside the AI core must plan for a world where a conventional Series B simply does not arrive.

Downstream: fewer Series A rounds, longer seed lives

The pressure does not stop at Series B. In the same late-October week, Crunchbase News reported Q3 2026 Series A counts falling to 272 from 377 the prior quarter - impressive headline totals masking a narrowing path for ordinary seed companies. Crunchbase's May 2026 analysis had already shown that the share of companies raising a $1M-plus seed and then progressing to a later round or an exit fell to 24% for the 2023 cohort and 16% for 2024, against above 55% for cohorts through 2020. Meanwhile the median seed round rose to about $3 million, triple the $1 million of five years ago, and the seed-to-A gap has exceeded two years since 2023.

Together those three numbers describe a portfolio-construction problem, not a mood. If seed cheques are larger, progression odds lower and waiting times longer, angel capital stays locked up for longer in companies less likely to be repriced by a later round. Returns do not disappear, but the channel shifts from mark-ups by downstream investors to self-generated cash flow or early acquisition.

Three questions to ask before signing a seed cheque today

First: how much capital does this company need to reach cash-flow breakeven without a Series B? Second: if a Series B happens, does it belong to the shrinking sub-$30M band or the expanding $100M-plus band, and why? Third: can today's deal structure survive a wait longer than two years, with at least one bridge round in between, without diluting founders past the point of motivation? Crunchbase data does not answer these, but it tells you the base rates have moved.