VC Funding Hits Record $510B in H1 2026, AI Dominates

Global startup investment just posted its biggest half-year on record — and a handful of AI companies swallowed most of it. According to data released by Crunchbase on July 2, venture funding worldwide reached $510 billion in the first half of 2026, more than the $440 billion invested across all of 2025 and the highest total for any six-month stretch on record.

The number is staggering on its own. The concentration behind it is the real story.

Two companies, 43% of the money

Of that $510 billion, OpenAI and Anthropic alone accounted for $217 billion — 43% of all global startup funding in the first half of the year. Two companies, out of the tens of thousands that raised capital worldwide, took nearly half the money.

The rounds behind those figures were historic in their own right. OpenAI finalized a $110 billion funding round at a $730 billion post-money valuation in May. Anthropic raised roughly $65 billion in the quarter and, on the back of it, became the most valuable private company on the Crunchbase Unicorn Board — the same Series H round that underpins Anthropic’s confidential IPO filing and drew a strategic memory stake from Micron.

For context, either of those single rounds would have been a record-shattering event a few years ago. In H1 2026 they were line items in a table.

The quarterly shape

The half split unevenly. Q1 2026 was the largest single quarter on record at $305 billion — inflated by the mega-rounds that closed early in the year. Q2 came in at $205 billion across more than 5,000 startups, a step down from Q1’s peak but still an enormous figure by any historical standard.

The AI tilt sharpened as the year went on. In Q2, more than 70% of all global startup capital went to AI-focused companies, up from just under 50% a year earlier. In the span of twelve months, AI went from about half of venture dollars to more than seven in ten. Every other sector — biotech, fintech, climate, enterprise software — is now competing for the shrinking remainder.

The exit window reopened

For most of 2024 and 2025, the venture industry’s complaint was liquidity: capital went in, but the IPO and acquisition markets stayed frozen, leaving investors with paper gains and no exits. H1 2026 broke that logjam.

Twenty-four companies were acquired at prices at or above $1 billion in Q2, totaling $113 billion in deal value — the highest quarter for large acquisitions on record. And the public markets cracked open at the top: SpaceX went public at a $1.77 trillion valuation, raising $75 billion in what ranks as the largest venture-backed IPO ever.

A reopened exit market matters for reasons beyond the headlines. Exits return cash to limited partners — the pension funds, endowments, and sovereign funds that supply venture capital its money. When that cash comes back, LPs recommit to new funds, and the cycle refills. A record fundraising half-year paired with a record exit quarter is the machine running at full speed in both directions at once.

The concentration risk nobody can ignore

The bullish read on these numbers is straightforward: capital is abundant, the exit window is open, and the technology absorbing the money is the most consequential in a generation. The bearish read starts from the same data and reaches a warier conclusion.

When 43% of global venture funding flows to two companies, and 70% of a quarter’s capital flows to a single sector, the entire asset class becomes a concentrated bet on one thesis: that AI’s revenue and productivity gains will eventually justify the capital being poured into models, data centers, and chips. If that thesis holds, today’s valuations will look cheap in hindsight. If it slips — if the unit economics of AI data centers prove harder than the spending assumes — the concentration cuts the other way, and a small number of very large marks drag the whole class down.

That tension isn’t abstract. It showed up in public markets the same week these figures landed, when AI chip stocks sold off sharply on exactly this worry — that infrastructure spending has outrun any near-term proof of returns. Private valuations don’t reprice daily the way stocks do, but they’re anchored to the same expectations. The largest private rounds in history are being written against a revenue future that is still, in large part, a forecast.

Below the mega-rounds

It’s worth separating two markets that the top-line number blends together. There is the mega-round market — OpenAI, Anthropic, xAI, a few neocloud infrastructure plays — where valuations and check sizes have detached from any prior reference point. And there is the everything-else market, the thousands of ordinary startups raising seed and Series A rounds, where conditions are far more normal and, for founders outside AI, arguably tougher than the record headline suggests.

Capital that concentrates at the top doesn’t automatically trickle down. A startup building payroll software or a climate-tech hardware company isn’t competing for the same dollars as OpenAI; it’s competing for the attention and allocation that’s left after the mega-rounds are funded. In a year when AI takes 70% of the money, the other 30% has to stretch across every non-AI ambition on earth.

What it means

The H1 2026 figures describe a venture market that is simultaneously the healthiest and the most lopsided it has ever been.

Winners: the frontier AI labs, which can now raise sums that would have been unthinkable and are effectively insulated from the capital constraints that discipline everyone else; their early investors and employees, for whom the reopened exit market — SpaceX’s IPO especially — turns paper wealth into realized returns; and the limited partners now receiving distributions after a long drought.

Losers, or at least the exposed: non-AI founders competing for a shrinking share of attention, and the venture asset class as a whole, which has quietly become a leveraged position on a single technological bet. Diversification is the first principle of managing risk, and by that measure the industry has rarely been less diversified.

What to watch next: whether Q3 sustains the pace or Q1’s record proves to be the peak; whether the Anthropic IPO and the queue behind it clear at their private valuations or force a public-market repricing; and, above all, whether AI revenue growth begins to visibly close the gap with AI spending. The money has already been committed. The half-year that decides whether it was wisely committed is still ahead. As always, the record that looks like validation on the way up is the same concentration that amplifies the fall if the thesis cracks.