Anthropic Just Surpassed OpenAI. Here's What Actually Happened.

On May 28, 2026, Anthropic announced a $65 billion Series H funding round at a $965 billion post-money valuation. That number puts it ahead of rival OpenAI, which closed a $122 billion round in March at an $852 billion valuation, according to TechCrunch.

Two private companies — neither of them profitable yet — are collectively valued at over $1.8 trillion.

Who's Writing the Checks

The round was co-led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. Institutional players including Baillie Gifford, Blackstone, Brookfield, D.E. Shaw Ventures, DST Global, and Fidelity also participated, per TechCrunch.

Strategic chip partners Samsung, SK Hynix, and Micron joined as well. $15 billion of the total came from previously committed investments — including a $5 billion pledge from Amazon announced back in April.

Altimeter Capital founder and CEO Brad Gerstner called Claude's adoption among enterprise customers the driver of this momentum. Anthropic CFO Krishna Rao said the funds will go toward compute expansion, safety research, and scaling Claude products.

The Valuation Tripled in Three Months

In February, Anthropic was valued at $380 billion, according to CNBC. By May 28, it hit $965 billion. That's a threefold increase in under four months.

The company is crediting this to its AI coding tool Claude Code and enterprise adoption. According to TechCrunch, Anthropic also unveiled Claude Opus 4.8 the same day as the funding announcement — a new flagship model with improved agentic coding capabilities, a 3x cheaper fast mode, and benchmark scores that beat GPT-5.5 across at least 12 categories, according to VentureBeat.

Timing a product launch with a funding announcement is standard practice for narrative management.

The $47 Billion Run Rate Nobody Should Accept At Face Value

AnthropIC says its revenue run rate crossed $47 billion earlier in May. CNBC confirmed the number. TechCrunch confirmed it. Bloomberg confirmed it. Almost nobody questioned it hard.

ZeroHedge did. And the skepticism isn't unfounded.

Analyst Jared Sleeper flagged on X that a single enterprise customer deal — reportedly $500 million per month — was being annualized to $6 billion in ARR. That one contract, extrapolated to twelve months, would represent a figure larger than Snowflake's entire run rate revenue as of Q1 2026. If Anthropic's headline number leans on a handful of deals like that, it is NOT the same as sustained recurring revenue.

Run rate metrics are legal. They're also a startup's favorite way to make hockey-stick growth look like a trend instead of a moment.

ZeroHedge argued the number is questionable and will be for the company's new public shareholders to evaluate in a few months.

That may be harsh. But the underlying math question is legitimate, and most mainstream outlets didn't even ask it.

What Claude Opus 4.8 Actually Is

Separate from the funding, the product news is solid.

According to VentureBeat, Opus 4.8 ships at unchanged API pricing — $5 per million input tokens, $25 per million output tokens — while slashing fast mode pricing by 3x compared to Opus 4.7. That makes high-throughput enterprise use cases significantly cheaper to run.

On benchmarks, ZDNET noted the model scores higher than Opus 4.7 on coding tasks but doesn't fully surpass GPT-5.5 across every metric. The publication also flagged that Anthropic is claiming Opus 4.8's alignment rates are similar to Claude Mythos Preview — a powerful cybersecurity model Anthropic has kept on a very tight leash due to safety concerns.

This raises questions about whether Anthropic has achieved a major safety milestone or whether the widely-released model is more capable than the company's safety approach suggests.

The IPO Race Nobody Is Pretending Isn't Happening

Both Anthropic and OpenAI are clearly racing toward public markets. TechCrunch described this Series H as potentially Anthropic's last private funding round before an IPO. Elon Musk's SpaceX — which merged with xAI — has already filed a prospectus with the SEC, per CNBC.

The Wall Street Journal reported that Anthropic expects a 130% revenue surge that would bring it to its first operating profit. That expectation carries significant weight. The company is NOT currently profitable.

Investors are not buying what Anthropic is today. They're betting on what it becomes after going public — and they need a valuation story to justify the price on the prospectus.

Conclusion

Anthropic is a serious company building serious technology. Claude Opus 4.8 is a real product with real enterprise adoption. The funding is real money from real institutions.

But a valuation that tripled in three months, a revenue run rate that quintupled in one year, and a number that may be built on annualizing short-term outlier deals raise serious questions.

When this thing goes public, retail investors will be the ones holding the bag if the math doesn't hold. The institutions writing nine-figure checks today have lawyers and exit strategies. Regular people buying shares on day one do not.

The AI boom is real. The valuation premium built on top of it could be substantial.