Loading today’s stories
Invest1 distinct publisher3 min readPublished
Crunchbase counts 58 billion-dollar venture listings in the first half against 27 a year earlier, but $86bn of the $110.8bn raised was SpaceX, which leaves the autumn calendar to prove the depth.
The Investor · Invest desk
Compiled by The InvestorSomething wrong?How this is made
The interesting number is the one left after the headliner. Take $110.8bn, subtract SpaceX's $86bn, and the other 57 companies that listed at a billion or more in the first half raised $24.8bn between them [1], about $435m each against $467m each for the 27 that listed in the first half of 2025 [3]. The count went up 2.15 times and the average raise came down roughly seven percent [7] [3]. SpaceX supplied 77.6% of the dollars [2], which makes the dollar column a description of one launch and the count a description of a market.
Against that, the Wall Street Journal's figure for Anthropic (up to $100bn, as soon as September or October) is about four times what the other 57 raised put together [4], and 16% more than SpaceX took [5]. It is also 0.8 times the $125bn Anthropic has already raised privately since 2021 [6], which is the tell about what the transaction does: the company has taken in more capital than the offering would provide, so the offering exists to move a very large private position into hands that can sell it [6].
Crunchbase's screen wants at least a 40% probability of a listing inside six months, and Anthropic does not clear it; the model prefers six to 12 months [4], while the reporting says weeks. Both readings survive, because slippage is nearly free for a company that does not need the money. Kraken filed confidentially with the SEC almost a year ago, paused amid volatility, said in May through chief executive Arjun Sethi that it was around 80% ready for a 2026 listing, and has reportedly weighed 2027 anyway [12]. Anthropic, for its part, has indicated it means to beat OpenAI out [7], and OpenAI on the same Journal account may push to 2027 [8].
The rest of the watchlist is where the window actually binds. Notion has raised $343m across its life, roughly a quarter of one percent of Anthropic's private total [8], and has seated a board with public-company experience in what Alex Konrad reported as a step towards an IPO [11], which buys optionality rather than settling anything. Oura raised $1.5bn and carries an $11bn mark from its last round [9], and its print would be the first serious read in years on what public money pays for consumer health hardware [10]. SambaNova at least has a comparable, Cerebras having priced a $6.4bn Nasdaq deal in May [13].
This is probably wrong, but I would rather have the count than the dollars: 58 listings at a billion or more [1] cannot be manufactured by a single issuer, whereas $110.8bn can be, and nearly was [2]. The counter-thesis, which I would not argue against hard, is that 69 companies listed in all of 2025 [1] and a backlog empties once, in which case the first half was catch-up and the second half thins out regardless of who files. What would prove me wrong is an autumn that delivers the mega-deals and not the count.
Ranked by verification strength, evidence, and original report placement.
In the first half of 2026, 58 venture-backed companies listed at $1 billion or above, per Crunchbase data, compared with 27 in the first half of 2025 and 69 in all of 2025.
Through the first half of 2026, venture-backed startups globally raised $110.8bn via IPO listings, versus $12.6bn in the first half of 2025, per Crunchbase.
SpaceX's June IPO launched it onto the Nasdaq and raised $86bn.
Crunchbase's predictive intelligence tools flag well-funded private companies with at least a 40% probability of going public within the next six months; Anthropic sits just outside that near-term screen, with the model favouring a six- to 12-month timeline.
Anthropic has raised $125bn from private-market investors since its founding in 2021, and its IPO would mark a major liquidity event for those backers.
OpenAI is deemed a very likely IPO candidate by Crunchbase but not within the next six months, and the WSJ report noted the company is considering pushing its listing to 2027.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · September 2, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
invest
Five years to $100B: the private market now finishes compounding before the IPO bell1 distinct publisher
leadership
The exit window reopened, but only for the very top of the AI stack2 distinct publishers
product
OpenAI's CFO calls an IPO "another fundraise" while the model calendar slips2 distinct publishers
build
The $559M-versus-$12.3B quarter matters more than the $65B run rate4 distinct publishers
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One dataset, one newsroom, two borrowed headlines
The 58 listings, the $110.8bn and the $86bn all come from Crunchbase's dataset, published by Crunchbase's newsroom, and nobody in this story checks them against a second tally. The two most consequential forward specifics — Anthropic's autumn window at up to $100bn, Oura's above-$11bn ambition — are summaries of a Wall Street Journal report we do not have in front of us. Only Kraken's SEC registration statement and Notion's board refresh, credited to Alex Konrad, have anything documentary or independently reported behind them.
Two deals priced, one filing paused, six maybes
Part of this is already banked: SpaceX priced in June for $86bn, Cerebras in May for $6.4bn, and 58 companies got billion-dollar listings away in six months. That is completed market activity, not a pitch deck. But of the eight names on the watchlist only Kraken has filed anything at all, and that filing has been sitting paused since volatility hit — the rest exist as model probabilities.
Record class, singular issuer
The framing is a record IPO year; the arithmetic is one rocket company. Take SpaceX out and the remaining 57 billion-dollar listings raised $24.8bn between them, and the average non-SpaceX deal — about $435m — came in below last year's roughly $467m. The count really did more than double, so this is not inflation across the board; it is a dollar headline and an eight-name pipeline both leaning on a single transaction and a probability screen rather than on filed paperwork.
The prediction is the product
Crunchbase's newsroom is reporting the output of Crunchbase's predictive intelligence subscription, and the piece takes care to explain the 40%-in-six-months threshold and then show it picking winners. None of that is concealed — it is stated in the second paragraph — but a story whose spine is the vendor's own model rewards confident calls and offers no way to check past ones. Worth noting that the two loudest numbers in the piece were reported by somebody else.
Firm on June, thin on October
One publisher, its own numbers, and a calendar mostly composed of things that have not happened. What is priced — SpaceX, Cerebras, the half-year tally — can be relied on as reported. Everything from September onward, including whether an up-to-$100bn Anthropic figure survives contact with a prospectus, rests on a secondhand report that Crunchbase's own model quietly disputes, and on a Kraken timeline that the same paragraph pushes both into 2026 and out to 2027.