Accel just raised $5 billion in new capital, and the numbers behind it tell you everything about where AI money is flowing right now.

The venture capital firm, which made early bets on Facebook and Slack before pivoting hard into AI, is deploying $4 billion through its fifth Leaders Fund and another $650 million through a sidecar vehicle. Average check size: $200 million. Target: 20 to 25 late-stage AI companies globally. Bloomberg first reported the raise.

Here is why this matters: Accel's existing AI portfolio is printing money at a rate that makes its Facebook returns look quaint.

The firm invested in Anthropic during its Series G at a $183 billion valuation. Anthropic has since closed a round at $380 billion and is now attracting offers around $800 billion. That is a quadruple in months. Its annualized revenue has hit $30 billion, a trajectory no company in history has matched.

Then there is Cursor. Accel backed the AI code editor in June 2025 at $9.9 billion. By November it had raised again at $29.3 billion. By March 2026, it was in discussions around $50 billion. That is a 5x return on a developer tool that barely existed two years ago.

The VC arms race has gone nuclear

Accel's $5 billion is enormous by any normal standard. In Q1 2026, it is almost table stakes.

The quarter saw $297 billion flow into startups worldwide, 2.5 times the total from Q4 2025 and the most venture funding ever recorded in a three-month period. Andreessen Horowitz raised $15 billion. Thrive Capital closed more than $10 billion. Founders Fund is finishing a $6 billion raise. The biggest funds are now measured in tens of billions, and single checks are measured in hundreds of millions.

This is not venture capital anymore. This is infrastructure finance. The firms raising these funds are not betting on product-market fit. They are betting on which companies will become the foundational layer of the economy for the next 30 years. And they are writing checks that would have funded entire fund vintages a decade ago.

The prepared mind, updated

Accel was founded in 1983 on what its founders called the "prepared mind" approach: deep sector research before investments materialize. The firm's most famous prepared-mind bet was its 2005 investment of $12.7 million for 10% of Facebook, a stake worth $6.6 billion at IPO.

The question is whether the same thesis works when checks are 200 million dollars and the companies you are buying into are already worth hundreds of billions. At that scale, you are not finding the next Facebook in a dorm room. You are picking winners in a market where the winners are already obvious and the only question is how much more expensive they get.

The broader portfolio tells you where Accel thinks the remaining value sits: Vercel (frontend deployment), n8n (AI automation), Recraft (design), and Code Metal (hardware development tools). These are the picks-and-shovels plays, the companies building infrastructure around the model layer rather than competing with it.

What happens next

Three IPOs are expected this year: SpaceX/xAI, OpenAI, and Anthropic. The combined target is roughly $240 billion in new public stock. Accel sits on the cap table of the one that is appreciating fastest. If Anthropic prices anywhere near $800 billion, Accel's Leaders Fund IV returns will be the stuff of Silicon Valley legend.

The uncomfortable truth for every other VC firm: the AI winners are already winning. Getting in now costs $200 million per ticket. Getting in late costs everything.