AI Venture Funding Q3 2026: Bigger Checks, Fewer Deals

A Quarter Carried by Its Biggest Deals

Third-quarter venture numbers look stable at first glance. Underneath, they show a market that is narrowing. Global venture investors deployed $145.6 billion across 2,756 disclosed-size rounds in Q3 2026. That was a 1.3% rise from Q2's $143.8 billion, but the number of rounds fell 10.1% from 3,067.1 Total capital held up while fewer companies received checks, and the average disclosed round reached $52.8 million.1

The concentration at the top explains the gap. Thirty-three rounds of $1 billion or more raised $62.8 billion, or 43.2% of the quarter's capital, up from 41.3% in Q2.1 Excluding those deals, the rest of the market raised $82.8 billion across 2,723 rounds, down 2.0% from the comparable Q2 figure.1 All of the quarter's growth came from mega-rounds. The rest of the market contracted slightly.

September's Rebound and a Shrinking Typical Deal

September was the quarter's strongest month by both dollars and deal count, with $53.1 billion across 1,012 rounds. Ten mega-rounds accounted for 39.4% of that total.1 The typical deal still got smaller. The median disclosed round fell to $5.55 million in September from $6.5 million in August.1

The falling median is the clearest sign of how unevenly capital is spread. Rising totals combined with a declining median suggest that a few very large financings are carrying the aggregate, while ordinary startups raise smaller amounts.

AI's Grip on Capital

Artificial intelligence remains the main destination for venture money. AI companies took 52.8% of all capital in September.1 In North America the share was higher still, with AI absorbing roughly two-thirds of the region's venture dollars in Q3.2 One analysis argues this leaves non-AI startups at the Series B and C stages competing for a shrinking slice of capital.2

Global Series B data puts that middle stage in context. Series B rounds raised $14.6 billion across 229 disclosed-size rounds in Q3, about 10.1% of the quarter's capital.1 When AI companies and billion-dollar rounds take most of the money, the growth-stage startups outside AI are the ones most likely to be squeezed.

Where the Numbers Diverge

The two datasets tell somewhat different stories, mainly because they measure different things.

The global count shows Q3 roughly flat to up, with 33 billion-dollar-plus rounds.1 A Crunchbase-based reading instead highlights a record 27 billion-dollar rounds globally. Its companion North America report shows the region's funding falling 35% quarter over quarter to $92 billion.2 The difference in mega-round counts likely reflects different methods and thresholds rather than a real contradiction. Both point to an unusually high number of very large deals.

The North America decline needs context. It was driven mainly by the absence of new OpenAI- and Anthropic-scale megarounds in the quarter, not by a broad retreat from the region.2 Those financings were so large in earlier quarters that any period without them shows a steep drop. Other figures support the idea that capital simply moved around. North America took 63.4% of global capital in September, and Europe had its strongest month in the tracked window, helped by Mistral AI's $3.49 billion Series D.1 Asian capital roughly halved.1

The Missing Exit Window

Exits are the weak point. Only 17 venture-backed companies went public in North America during the quarter, raising under $4 billion combined.2 This matters because current AI valuations, especially for leaders such as OpenAI and Anthropic, assume investors will eventually get liquidity. That exit window has not yet opened.2

The Takeaway

The main lesson from Q3 is concentration, not growth. Headline totals depend on a few very large rounds, mostly in AI. Excluding them, the market shrank slightly. Deal counts and the median check also fell.

The headline figures should be read with caution. A quarter can look strong or weak depending on whether one frontier lab happened to close a round. North America's 35% drop and the global 1.3% gain can both be accurate at once.12

The risk is a market with heavy money flowing in and little flowing out. If IPOs stay scarce while capital keeps concentrating in a small group of AI companies, smaller startups outside that group will face tougher fundraising. The broader ecosystem will also stay dependent on liquidity events that have not yet happened.