$65 Billion in Fresh Funding

Anthropic closed a $65 billion Series H round on May 28, 2026, at a $965 billion post-money valuation, according to TechCrunch. A company that didn't exist before 2021 is now worth nearly one trillion dollars.

The round was co-led by Altimeter Capital, Dragoneer, Greenoaks, Sequoia Capital, Capital Group, Coatue, and D1 Capital Partners. Institutional names like Blackstone, Brookfield, Fidelity, and D.E. Shaw also joined. So did semiconductor players Samsung, SK Hynix, and Micron — chip companies buying a seat at the table of their biggest future customer.

$15 billion of that total comes from previously committed investments by hyperscalers — including the $5 billion Amazon pledged in April.

Altimeter Capital founder Brad Gerstner said: "Claude's latest advancements have driven large-scale adoption among the world's most demanding organizations."

The Numbers Behind the Valuation

Anthropic says its annualized run rate revenue crossed $47 billion earlier this month. The Wall Street Journal reported the company projects a 130% revenue surge that would take it to its first operating profit.

For context: OpenAI raised $122 billion in March 2026 at an $852 billion valuation. Anthropic just lapped them on the valuation scoreboard while raising roughly half as much capital. Either Anthropic's investors see something OpenAI's don't, or the AI funding market has lost any remaining contact with reality.

Opus 4.8: A Real Upgrade, 41 Days After the Last One

The fundraise dropped the same day Anthropic released Claude Opus 4.8 — a model that arrived just 41 days after Opus 4.7. That's an unusually fast turnaround for Anthropic.

Opus 4.7 received mixed reviews from users. Google's Gemini Flash and OpenAI's Codex both pushed out notable releases in that same window, putting Anthropic on defense.

Opus 4.8's headline feature is honesty. Anthropic claims the model is "more likely to flag uncertainties about its work and less likely to make unsupported claims." According to Anthropic's own evaluations, Opus 4.8 is roughly 4x less likely than its predecessor to let code flaws pass without comment.

Bridgewater Associates tested the model and told Anthropic the biggest improvement was "Opus 4.8's tendency to proactively flag issues with the inputs and outputs of an analysis, something other models routinely missed."

Anthropic also launched Dynamic Workflows in research preview. The system lets Claude Opus 4.8 spin up hundreds of parallel subagents in a single session to tackle large-scale tasks — think full codebase migrations across hundreds of thousands of lines of code. Users can also dial Claude's effort level up or down to manage token usage.

The Mythos Model Is Still Locked Up

Anthropic's Mythos model — its most powerful, with significant cybersecurity capabilities — remains in limited release. The company hinted in the Opus 4.8 launch post that Mythos could reach general availability "in the coming weeks" once safety safeguards are in place, according to TechCrunch.

The SpaceX Deal Is Still a Mess — And It Got More Complicated

Elon Musk went on X and directly contradicted SpaceX's own SEC filing.

Musk posted that the Anthropic compute deal is a "180-day lease with 90-day notice mutual cancellation" — framing it as short-term and flexible. He said the short term was SpaceX's choice, not Anthropic's.

That directly conflicts with SpaceX's S-1 filing, which TechCrunch reviewed in detail. Page F-62 states the deal runs through May 2029 — a three-year agreement. Pages 13 and 146 of the filing go further, explicitly stating Anthropic "has agreed to pay $1.25 billion per month through May 2029." The same language appears on pages F-96, 13, and 146. This was NOT a one-time mention buried in a footnote.

xAI did NOT respond to TechCrunch's request for clarification. Neither SpaceX nor Anthropic has publicly addressed the discrepancy.

SpaceX is in a quiet period ahead of its IPO. Making false or misleading statements to investors during a quiet period is exactly the kind of thing the SEC exists to penalize. TechCrunch noted this directly — and added that the SEC "probably will not" act on it.

What Mainstream Coverage Is Missing

Every outlet is treating the $65 billion raise as a triumphant AI moment. But almost no one is asking: who actually owns Anthropic at a $965 billion valuation, and what are they buying?

Amazon is already in deep. Now Samsung, SK Hynix, and Micron — Anthropic's hardware supply chain — are also investors. That's vertical integration in slow motion. The company claiming to lead on AI safety is increasingly owned by the infrastructure companies that profit most from AI proliferation.

And the SpaceX filing contradiction needs serious attention. Musk's X post directly conflicts with sworn SEC language. Someone is wrong. The public doesn't know who. Neither company is talking.

What Comes Next

Anthropic is legitimately impressive — real revenue, real enterprise adoption, a model that major institutional investors are calling a meaningful upgrade. The $965 billion valuation may be frothy, but it's not imaginary.

But the SpaceX deal discrepancy is unresolved. Either Musk misrepresented a legal contract to the public, or SpaceX's SEC filing is materially wrong. Both are serious problems.

For regular people: if Anthropic goes public anywhere near $1 trillion, your retirement fund will probably end up owning a piece. Honest answers about what's in those contracts matter.