Anthropic Raises $30 Billion at a $380 Billion Valuation as Claude Code Passes $2.5B Run Rate
The Series G is among the largest private rounds ever, and it values Anthropic at more than ten times where it stood a year earlier.
Key takeaways
- Anthropic closed a $30 billion Series G at a $380 billion post-money valuation, led by D. E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ and MGX.
- Run-rate revenue is about $14 billion, and Claude Code alone is above $2.5 billion; the company is not yet profitable.
- Later reports put a Series H at $965 billion in mid-2026, though that figure rests on secondary sources.
Anthropic announced a $30 billion Series G on February 12, 2026, valuing the company at $380 billion after the money went in. The round, led by D. E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ and MGX, arrived weeks after a smaller raise at $350 billion and underlined how quickly investors are repricing the frontier labs.
The round in numbers
The $30 billion makes this one of the largest private financings in history. Anthropic said it will use the money for research, product development and infrastructure expansion, and for making Claude available to more customers around the world. The company also disclosed a revenue figure: run-rate revenue of roughly $14 billion, which it says has grown about tenfold in each of the past three years. Claude Code, the coding agent, accounts for more than $2.5 billion of that on its own. Anthropic has not yet reached profitability.
The round followed an earlier January close at a $350 billion valuation that exceeded its $10 billion target, reported by Benzinga. Two raises within weeks of each other are a sign that demand from investors outran the company's initial plan, and that Anthropic was willing to take more capital rather than leave it on the table.
Where the money comes from, and goes
The mix of backers matters. D. E. Shaw and Dragoneer are crossover investors that normally buy into late-stage and public companies; Founders Fund and ICONIQ are long-standing technology backers; MGX is an Abu Dhabi investment vehicle focused on AI. That blend reads like a pre-IPO syndicate, and reports through the spring that Anthropic was weighing a public listing fit the pattern, although those reports remain unconfirmed.
Why the number matters
A $14 billion run rate against a $380 billion valuation is a multiple of roughly 27 times, steep for a software company but modest next to the growth rate Anthropic claims. The more important point is the shape of the revenue. Claude Code's $2.5 billion shows that a single agentic product aimed at developers can become a major business within months, and it explains why coding benchmarks dominate model launches. Every point gained on SWE-bench is a sales argument to engineering teams.
The funding also shows the financial arms race. Training frontier models and serving them at scale requires data centres and chips measured in gigawatts and billions of dollars, and the labs are not profitable. OpenAI closed a $122 billion round at $852 billion on March 31, which The Next Web reported, so the two leaders are raising capital at a pace few companies have ever managed. Later reports suggested Anthropic raised again, a $65 billion Series H at $965 billion around June 1, which would put it ahead of OpenAI; that rests on secondary coverage and should be checked against a primary announcement.
The competitive read-through
Rival labs will study the disclosure as closely as investors do. A claimed tenfold annual growth rate for three years running shows how new this market is, and it sets a bar: if Anthropic sustains even half that pace, it would pass $25 billion in run-rate within a year, which is the scale at which a public listing becomes realistic. Enterprise customers, who buy through the API and through cloud marketplaces, are the main driver, and coding agents are the product category pulling them in.
The risks are mirrored. Heavy dependence on a few model generations means each release must land, and each rival launch, from Gemini or GPT, can erode share within weeks. Compute commitments are fixed costs that run ahead of revenue, so a slowdown in demand would hurt fast. Investors are pricing in continued dominance of coding and agent workloads, which is a narrower thesis than "AI in general".
What to watch
- Whether Anthropic files for an IPO and what audited financials reveal about margins and compute commitments.
- Concentration risk: how much run-rate depends on coding customers and on a handful of large enterprise accounts.
- The valuation gap with OpenAI, and whether the Series H figure is confirmed by Anthropic.
- How much of the raise flows to chips and data centres versus research salaries.
