Back in 2024, Pete Martin found himself closing a $5 million seed investment for his AI-powered cybersecurity project, Realm. At the time, the $25 million post-money valuation felt like a daring leap—an audacious number for a young company just getting off the ground. But if you fast-forward to today, Martin’s story now sounds almost quaint. In this climate, seed rounds of $10 million with post-money valuations north of $40 million are routine, especially if AI is at the heart of your company’s pitch.
The truth is, such numbers are reserved almost exclusively for AI upstarts: investors have all but turned their backs on sectors that aren’t riding the AI wave. Ashley Smith, a general partner at Vermilion, remembers the electricity in the air during the March Y Combinator Demo Day. Attendees murmured—sometimes loudly—about how inflated the company valuations had grown. She saw multiple startups, some barely two months old, waving six- and seven-figure customer contracts. One newcomer, astonishingly, was not just asking but expecting to raise $5 million at a $40 million valuation.
Smith didn’t buy the idea that this was some special premium reserved for Y Combinator companies. It wasn’t the legendary “YC markup.” The truth felt simpler: Round after round, investors were valuing companies years ahead of their actual milestones, hungry for first dibs on whatever could turn out to be the next legendary tech rocket.
Flush with fresh capital, the heavyweight venture firms have started to insert themselves into the earliest funding rounds, further inflating prices. Even mid-sized funds, Smith says, are crowding in, their appetite for AI seemingly insatiable. For VCs focusing on infrastructure—Smith among them—it’s easy to get nudged out when the big players swoop in. That’s why, although the number of seed deals has shrunk, the valuations on those that get done have soared. The data backs up what everyone is feeling.
Some, like Shanea Leven of AI app platform Empromptu, lay the blame at the feet of breakout stories like Cursor. In the first 12 months of 2025, Cursor rocketed to $100 million in revenue, turning the industry’s expectations upside down. Similar feats by Lovable, Bolt, OpenEvidence, and ElevenLabs set a blistering pace, though few can really keep up. Yet, even founders who aren’t outliers describe feeling the heat radiating through the sector.
Pressure has become the new norm. Investors, Leven points out, aren’t looking for the first $1 billion company these days—the bar is now a wild $50 billion. With so much capital swirling around this space, everyone’s breathless pace is the only way not to be trampled underfoot.
Venture capitalists try to rationalize the madness. Marlon Nichols, the managing GP at MaC Ventures, tracks the changes by the size of the checks he’s writing: in 2019, a $1 million commitment was normal. Now, first checks hit $2.5 million or more, sometimes going as high as $5 million for companies still in their infancy. “The best seed companies now look nothing like seeds used to,” Nichols says. Advanced AI tools allow founders to move at breakneck speed from idea to paying enterprise clients. His last two investments were drawing in more than $2 million in revenue almost from day one, getting million-dollar checks and high valuations because the early momentum was impossible to argue with.
But traction is only part of the equation. Investors want founders who bring unmatched experience—graduates of OpenAI or veterans of other AI giants—since talent in this field is a precious, extremely expensive commodity. Amber Atherton from Patron calls it a “war for researchers.” In this race, backgrounds and track records count, sometimes as much as revenue.
This jockeying pushed seed valuations into the stratosphere. Thinking Machine Labs, led by ex-OpenAI executive Mira Murati, famously secured a $2 billion seed at a $12 billion valuation. Leven herself, a second-time founder, says her company now commands a valuation twice as high as her first venture did at a similar stage—and with far more real-world results to show. “Right now,” she says, “if you don’t already have multiple six-figure contracts lined up, you’re invisible to investors.”
As seed rounds inch out of reach, many VCs are angling earlier, championing “pre-seed” as the new seed. Smith and others see it as rolling the clock back to a time when pre-revenue was the norm for investing. Jonathan Lehr of Work-Bench says his firm, with a $160 million fund, has grown boldly comfortable with betting on teams pre-traction—just to stay ahead. The logic is simple: get onboard early, or risk missing the breakout story entirely.
Even so, the calculus isn’t simple. The expectations for young companies have ballooned. Atherton puts it plainly: launching and shipping a product is no longer enough. “Now it’s about telling a convincing story about how you’ll out-execute everyone else. That’s what gets a round closed.” The crushing weight of these raised standards leaves founders with little margin for trial and error.
The upside? More cash buys speed and firepower—key for racing against giants and running expensive AI systems. The downside is stark: If you lag on growth, you’re left holding the bag on a high valuation nobody wants to match. Lesson learned, Pete Martin warns: “You risk getting stuck in limbo—too pricey for next-round investors, but not enough results to prove your worth.”
It’s a new era. Faster. Riskier. And for some, thrillingly high-stakes—until the music stops.
