Series A Is Now $48 Million: Startup Funding Is Broken
The median Series A valuation hit $48 million in early 2025. Ten years ago, Series A meant raising $5
The median Series A valuation hit $48 million in early 2025. Ten years ago, Series A meant raising $5 million to $15 million to prove your business worked. Now it’s $18 million rounds at $48 million valuations, with over 40% of all Series A funding going to rounds of $100 million or more. Safe Superintelligence raised $1 billion. Merge Labs raised $252 million in a seed round before even getting to Series A.
Only 36% of companies that raised seed in 2021 made it past seed. For 2022, it’s 20%. Previous years saw 51% to 61% success rates. The gap between seed and Series A has become a death valley. First-time founders are frozen out whilst serial entrepreneurs raise nine-figure rounds on PowerPoint decks.
The Mega-Rounds
In Q1 2026, over 40% of seed and Series A money went to rounds of $100 million or more. A small number of founders with track records raise enormous sums. Everyone else fights over scraps.
Safe Superintelligence raised $1 billion in Series A. No product, no revenue, barely any employees. Ilya Sutskever’s reputation and AI hype were enough. Mal, an Islamic banking startup, raised $230 million in its first round. Upscale AI took $200 million. These aren’t outliers anymore. This is how Series A works if you’re in the winner’s circle.
Merge Labs raised $252 million in what they called seed. Sam Altman’s involvement helped. The round used multiple tiers with different valuations depending on liquidity preferences. The “headline” number didn’t reflect what most investors paid. But the message was clear: with the right backers, you can raise Series A money before Series A.
The number of jumbo deals grew significantly in 2024 and 2025. Investors decided spreading capital across many bets was less profitable than concentrating in a few winners. The “spray and pray” model died. Now it’s “bet big on proven founders.”
The Crunch
Seed rounds plunged 28% year-over-year in Q1 2025. Series A fell 10%. Fewer companies make it to the starting line. The explosion of micro-VCs and accelerators created thousands more funded startups, but legitimate Series A investors haven’t grown proportionally. The ratio of seed to Series A deals went from 1:1 in 2008 to roughly 2:1 by 2011. It’s only accelerated since.
Over 1,000 startups get orphaned each year, stranded between seed and Series A. The median time between seed and Series A stretched to 616 days, roughly 20 months. The time between Series A and Series B hit 28 months in 2024, longest since 2012. Fundraising takes longer whilst burn rates stay high.
The metrics that got you seed funding won’t get you Series A. The old threshold of $1 million annual recurring revenue doesn’t work. Investors want $3 million to $5 million ARR minimum, efficient growth, and a clear path to $50 million revenue within three years. For consumer companies without recurring revenue, the bar is even higher.
Of over 4,400 US companies that secured Series A in 2020 or 2021, just over 1,600 raised a Series B. That’s 36%. Many raised exceptionally large sums during the 2021 boom, giving them runway to extend time between rounds. Eventually the money runs out.
Only Proven Founders Win
The market decided backing first-time founders at early stages is too risky. Better to wait until they prove themselves, then back them heavily on company number two or three. How do first-time founders prove themselves if they can’t raise capital? They mostly don’t.
AI startups get valuations 42% higher than non-AI peers at seed. At Series A, median AI valuations exceed $50 million.Series B reaches $143 million. The premium goes almost exclusively to founders with track records or credentials from top AI labs.
Ilya Sutskever got $1 billion because he co-founded OpenAI. Sam Altman backs seed rounds that raise $250 million because he runs OpenAI. Former executives from Google DeepMind and Anthropic raise massive rounds based on where they worked, not what they’ve built. The credential premium is enormous.
Fintech, healthcare, infrastructure startups follow similar patterns. Investors want founders who’ve seen hyper-growth, who understand scaling from $10 million to $100 million revenue, who’ve built teams and navigated regulations. First-time founders haven’t done that.
The few exceptions prove the rule. Stanford PhD working on something deeply technical? Maybe. Unfair distribution advantage through connections to big companies? Possibly. Solving a problem for an industry you worked in for 15 years? Perhaps. Generic B2B SaaS? Forget it unless you’ve already exited.
The New Funding Math
The median Series A round in Q3 2025 was $18.1 million, though this varies significantly by sector. The median pre-money valuation was $49.3 million for primary rounds and $45.9 million for bridge rounds. Dilution in Q1 2025 averaged 17.9%, down from 20.9% a year earlier. Founders who demonstrate capital efficiency keep more of their companies whilst raising similar amounts.
But these medians hide the distribution. The average is pulled up dramatically by mega-rounds. Most Series A deals are either $5 million to $10 million for companies barely scraping by, or $100 million-plus for hot AI startups. The traditional $15 million to $25 million Series A for a solidly growing B2B company is disappearing.
Investors use revenue multiples of 3x to 10x annual recurring revenue to value Series A companies, depending on growth rate and market dynamics. A company with $5 million ARR growing 200% year-over-year might get a 10x multiple, implying a $50 million valuation. A company with $3 million ARR growing 80% might get a 5x multiple, implying $15 million. But these are just frameworks. Mega-rounds ignore the math entirely.
Companies that closed Series A rounds in the first half of 2024 had an average of 15.6 employees, 16.3% lower than Series A companies five years ago. Successful companies are doing more with less. The days of raising $20 million to hire 50 people and figure things out are over. You need to show product-market fit, efficient customer acquisition, and a repeatable sales process before Series A.
The Reality for First-Timers
You’re a first-time founder without AI credentials or a Stanford PhD. You raise $500,000 to $2 million in seed from angels or micro-VCs. You build a product, get customers, reach $500,000 to $1 million in annual revenue. Then you start looking for Series A.
You pitch 50 to 100 VCs over six months. Most don’t respond. Some take meetings out of politeness. A few say they’re interested but want more traction. None commit. The feedback is vague. “Come back at $3 million ARR.” “We need more efficient growth.” “The market timing doesn’t feel right.”
You extend runway by cutting costs, maybe raise a small bridge from existing investors. You hit $1.5 million ARR after 18 months. Still not enough. Burn rate means nine months of runway left. You go back to investors. Same response. The goalpost moved. Now they want $5 million ARR.
You run out of time. Options are acquisition, shutting down, or pivoting to something VCs care about. Most choose acquisition, selling for $10 million to $30 million if lucky. Founders and early employees make some money. Investors get 2x to 3x. Nobody gets rich. The company never reached its potential because it couldn’t access capital.
This pattern repeats thousands of times per year. Good companies with real customers and revenue die. Not because the businesses are bad, but because they’re not in the mega-round category.
Why VCs Went All In
Venture returns are highly concentrated. A few massive wins generate all the profits. The rest barely matter. If you’re a VC, your incentive is finding the 0.1% of companies worth $10 billion-plus. Everything else is noise.
The data supports concentrating capital. Spreading $100 million across 20 companies at $5 million each gives you 20 shots. Putting $100 million into one company gives you one shot, but if it works, you own a much larger percentage. The maths favours concentration when you believe you can identify winners early.
AI reinforced this. If AI creates trillion-dollar companies, the winners will be worth 10x or 100x previous tech giants. Missing the next OpenAI would be catastrophic for returns. Better to pay up and make sure you’re in. This drives valuations up and makes VCs less willing to bet on unproven founders.
VCs shifted from portfolio diversification to concentrated bets. They write fewer, larger cheques to companies they’re highly confident in. This leaves less capital for everyone else.
The Death of the Middle
Series A worked when it was $10 million rounds for companies that proved product-market fit and needed capital to scale. It doesn’t work when it’s $200 million rounds for PowerPoint decks from famous founders competing with $8 million rounds for companies generating real revenue. The traditional middle ground disappeared.
You’re either in the winner’s circle raising mega-rounds, or you’re fighting for scraps that barely move the needle. The funding model that built the startup ecosystem for 30 years is dead. What replaces it isn’t clear yet. Probably more concentration, more mega-rounds, more first-time founders shut out.
The winners will win bigger. Everyone else will struggle to survive. That’s Series A funding in 2026.
