Anthropic Hits $965B Valuation, Overtakes OpenAI as World's Most Valuable AI Startup

Three months is a long time in AI. In February 2026, Anthropic closed a Series G at a $380 billion valuation — itself a staggering figure that would have seemed impossible two years prior. By May 28, 2026, the company had more than doubled that figure, closing a $65 billion Series H round at a $965 billion post-money valuation. The result: Anthropic is now the most valuable private company in the AI industry, leapfrogging OpenAI — which filed its S-1 for a public offering valuing it at $852 billion — to claim a title that, until recently, seemed locked up by the maker of ChatGPT.

The round was led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. But the most telling aspect of the investor roster is who else participated: Micron Technology, Samsung, and SK Hynix — three of the world’s largest semiconductor manufacturers — all took positions. That semiconductor giants are writing equity checks into an AI model lab signals something fundamental about how the AI supply chain is evolving. Securing favored-customer relationships, or simply insurance, with the companies most likely to consume their cutting-edge memory and compute is now a board-level priority.

Claude Code: The Revenue Engine

Anthropic’s $47 billion annual revenue run rate — crossed in May 2026 — did not emerge from consumer subscriptions or chatbot usage alone. The dominant driver has been Claude Code, Anthropic’s AI coding assistant, which has become deeply embedded in enterprise software development workflows. As AI-generated code has shifted from novelty to standard practice, companies have adopted Claude Code at the infrastructure layer rather than the tooling layer — meaning it touches every developer, every pipeline, and every deployment.

Enterprise adoption of Claude across industries compounds the picture. Financial services firms use Claude to automate document review and compliance workflows. Healthcare organizations deploy it for clinical note summarization. The pattern is consistent: Claude enters through a pilot, proves productivity gains in weeks, and becomes load-bearing infrastructure within months.

The $47B run rate, crossing just three months after the $380B Series G, represents extraordinary velocity. For context, OpenAI’s S-1 filing, submitted in June, disclosed its own revenue trajectory — but Anthropic’s figure, if sustained, would put it among the fastest-growing enterprise software businesses in history.

Amazon’s Decade-Long Bet

Amazon’s commitment to Anthropic, disclosed alongside the Series H, deserves its own paragraph. Amazon has committed up to $25 billion in investment, and Anthropic has in turn pledged to spend over $100 billion on AWS cloud infrastructure over the next decade. That is not a vendor relationship. That is a structural coupling — the kind that typically shows up in the history books of how major technology platforms were built.

The arrangement solves a genuine problem for Anthropic. Training and serving frontier AI models requires compute at a scale that taxes even well-funded organizations. By locking in long-term capacity on AWS, Anthropic trades some financial flexibility for predictable infrastructure access. For Amazon, it provides an anchor tenant for its rapidly expanding AI-optimized data center footprint, and a model that competes directly with OpenAI’s models on Azure.

The OpenAI Contrast

The timing of Anthropic’s valuation announcement — arriving weeks before OpenAI’s S-1 filing — is not coincidence. The two companies have been on parallel trajectories since Dario Amodei and a core team of safety-focused researchers departed OpenAI in 2021 to found Anthropic. For years, the narrative held that OpenAI was the industry leader and Anthropic the principled challenger.

That narrative is now in flux. OpenAI’s public filing targets an $852 billion IPO valuation, below Anthropic’s latest private mark. More importantly, OpenAI’s path to public markets brings scrutiny that private companies avoid: quarterly disclosures, public earnings calls, and institutional shareholders who can vote. Anthropic remains private, with the flexibility that entails.

The strategic divergence between the two companies is sharpening. OpenAI has pursued consumer reach aggressively, deploying ChatGPT across voice interfaces, operating systems integrations, and social media. Anthropic has been more deliberate, investing heavily in interpretability research and enterprise reliability. Neither strategy is obviously correct — the market will decide.

The Capacity Problem

Despite the capital, Anthropic faces a problem that would be enviable in most industries and is merely stressful in this one: demand exceeding supply. The company has implemented usage limits during peak hours, forcing developers and enterprise customers to shift workloads to off-peak windows. Infrastructure provisioning at this scale — particularly for inference, where latency is customer-visible — is genuinely hard, and Anthropic’s operational build-out has lagged its commercial success.

The $100 billion AWS commitment is partly a solution to this problem. Deploying that capital takes time, but the runway is now substantial. Anthropic has also been expanding its own infrastructure team, recruiting from hyperscalers, and investing in custom silicon partnerships — the chipmaker investor syndicate is not unrelated to this effort.

What $965 Billion Means

To put the number in perspective: $965 billion is larger than the current market capitalization of most Fortune 500 companies. It positions Anthropic just below the trillion-dollar threshold that has only been breached by a handful of technology companies, all of which spent decades building to that point. Anthropic is roughly five years old.

The valuation reflects several converging beliefs among sophisticated investors: that AI model providers will capture a substantial share of enterprise software spending; that safety-focused development is commercially competitive, not a drag; and that Claude’s architecture and Constitutional AI training approach give Anthropic durable differentiation in a market that could otherwise commoditize.

Whether those beliefs prove correct depends on execution over the next few years — particularly on whether Anthropic can scale infrastructure to meet demand, maintain model quality advantages as competitors close gaps, and navigate the regulatory environment that is tightening around AI systems on both sides of the Atlantic.

What is not in doubt is that the competition at the frontier of AI is now a genuine two-company race, with a supporting cast of challengers. The prize — controlling the infrastructure layer that increasingly runs the global economy’s knowledge work — is large enough to justify nine-digit bets. Anthropic has just raised the stakes.