Which VC raises next? Managers holding top AI names took 63% of Q3 venture fundraising
What's the story? Which VCs are about to raise again? Murph CapitalDealroom has a profile for this one. Try Dealroom →, the emerging-manager research and media platform founded by Pavel Prata, tested whether owning a piece of the hottest AI companies makes the next fundraise easier, using its own Q3 fundraising data and startup data from Harmonic. It does.
The numbers: In Q3 2026, managers holding at least one of the 25 hottest AI companies closed 26 of 152 funds (17%) but took $28.8 billion, or 63%, of the $46.0 billion raised. Of the 24 such managers Murph could size, 21 raised more than last time, a median of 1.3x. Across a wider set of 429 US VCs that hold at least three of 1,889 venture-backed AI startups, managers with no top-25 stake filed 43% fewer new funds in 2025-26 than in 2023-24, while those with four or more stakes filed 35% more. After controlling for portfolio size and past activity, holding even one top-25 stake nearly doubled the odds of getting a new fund done (1.8x). Only early stakes counted: positions bought in 2025 or later made no measurable difference.
Speed: Managers with no top-25 stake took a median 34 months to file their next fund after a 2023-24 raise, against 23 months with two or three stakes and 20 months with four or more.
Who's next? From Form D filings going back to 2009 (215 raises by 26 managers, typical gap 18 months), Murph counts 11 of those 26 already back in market. Its calls: Dragoneer files next, likely in January; Lightspeed by spring; Thrive Capital by mid-2027, and bigger; Accel's next rotation around August 2027; a16z's next platform raise around end-2027; Founders Fund is the wildcard (2027 or late 2028); and Benchmark, on a three-year cycle, not until mid-2029. These are forecasts, not announced raises.
Source: Pavel Prata on X (27 October 2026)
