Skip to content
Vietnam Angel Network
Deep analysisAngel investment27 August 20267 min readAuto-researched

US angel capital 2026: bigger cheques, fewer companies funded

The Angel Capital Association's 2026 Angel Funders Report shows angel dollars up 12% but concentrated in fewer companies, at a time when the share of seed-backed startups reaching a next round has fallen sharply.

US$491.3M

Angel investment reported by ACA member groups for 2025

+12%

Increase from US$437M in 2024

≈2/3

Share of reporting angel groups with at least one AI-related investment

Global startup funding is at an all-time high. Crunchbase counted US$515 billion of venture capital invested in the first half of 2026 — more than the US$440 billion recorded for all of 2025 — and US$65 billion in July 2026 alone, with 14 rounds of US$1 billion or more, the highest monthly count of billion-dollar rounds on record. The earliest layer of capital, money from angel investors, is moving on a different logic entirely.

According to the 2026 Angel Funders Report published by the Angel Capital Association (ACA) on 13 July 2026, angel investment reported by member groups rose 12% year over year, from US$437 million in 2024 to US$491.3 million in 2025. The ACA describes this as the first sign of a disciplined recovery after several years of correction. What matters is how the money was distributed: groups deployed more capital per investment and per company — greater selectivity, heavier concentration.

More money, fewer doors

A market where total dollars rise while the number of funded companies falls is a market that favours the supplier of capital, not the seeker. For founders, the evidence bar has moved: a deck and a prototype no longer suffice. Angel groups want paying customers, a computable cost of customer acquisition, and a clear reason why angel money is the right money now. For investors, portfolio quality improves but the number of bets falls, and that changes the risk arithmetic: fewer positions mean wider dispersion of outcomes.

Scale matters here. All angel investment tallied by the ACA for 2025 is far smaller than a single billion-dollar round closed in July 2026. Angels do not compete on price with mega-funds, and should not try. Their edge sits at the stage institutional capital cannot yet reach: the pre-seed round, where decisions rest on sector knowledge and direct contact with founders.

Applied AI, not foundation models

The ACA reports that nearly two-thirds of contributing angel groups made at least one AI-related investment during the year, favouring applied AI, healthcare AI and industry-specific solutions over foundation models. Together with life sciences, these are the clearest destinations for capital in the association's 2026 data.

Structurally, that choice makes sense. Foundation-model labs raise at a scale no angel can follow, and Crunchbase data shows billion-dollar rounds at record frequency. By contrast, a company selling AI software to clinics, factories or logistics operators can reach real revenue on a few hundred thousand dollars of initial capital, and its value is tied to operating data and customer relationships — two things local angels judge better than foreign funds.

The follow-on arithmetic has changed

A Crunchbase analysis published in May 2026 (by Gené Teare) shows the median US seed round at roughly US$3 million in 2025, about three times the 2018 level, with an upper-quartile median of US$5.6 million. Graduation, however, has deteriorated: before 2020, companies raising a seed round of US$1 million or more typically progressed to a further round or an exit at rates of about 55% or higher; of the 2023 cohort, only 24% had progressed as of the report.

Put together, those two figures produce an uncomfortable result for angels: higher entry prices, longer holding periods and a lower probability of a subsequent round that reprices the position. It explains why ACA groups are writing more per company: if only a quarter of a portfolio will ever be repriced by the market, owning a meaningful stake in high-conviction names matters more than spreading thin.

  • Reserves are no longer optional: capital for follow-on should be earmarked at the moment of the first cheque.
  • Assume holding periods of five to seven years, not three, when only about a quarter of the 2023 US seed cohort has progressed.
  • Design the exit route up front — M&A and secondary transfers — rather than defaulting to a Series A.
  • Group-based due diligence lowers cost per investor and raises selection quality, the direction ACA groups are already taking.