Summary: Anthropic closed a $30 billion Series G on February 12, 2026 at a $380 billion post-money valuation, led by GIC and Coatue. The announcement also said more than 500 business customers each spend over $1 million annually, showing Claude’s enterprise use had moved from pilots into standing operating costs.
Anthropic confirmed on February 12, 2026 that it has closed a $30 billion Series G, landing at a $380 billion post-money valuation with GIC and Coatue leading the round. CNBC and TechCrunch framed the deal the same way in their coverage: the second-largest private round on record.
Tucked into the same announcement was a figure that deserves a closer look: more than 500 business customers each spend over $1 million a year with Anthropic.
Sizing the Second-Largest Private Round Ever
A $30 billion raise sits well outside the range of an ordinary growth round — this is capital deployed at infrastructure scale. Set it against the rest of February’s news and the context sharpens: SpaceX wrapped its merger with xAI on February 2 at a combined valuation of roughly $1.25 trillion, and reports on February 4 placed Amazon in advanced talks to invest as much as $50 billion in OpenAI. At the top of the AI market, capital operations now clear sovereign-fund and trillion-dollar-club levels, and this round stacks Anthropic’s chips on the same table.
Who led the round carries as much information as its size. GIC and Coatue did not come aboard as strategic cloud partners shopping for distribution; they are institutional money placing a direct bet on model-company growth. Anthropic has now taken capital from investors whose return case only pays off if enterprise AI spend keeps compounding — a wager on the category rather than on any single deal.
The Signal in 500 Million-Dollar Accounts
The hardest leap in enterprise AI adoption is the one from pilot budget to standing operating cost. More than 500 enterprises each spending a million dollars a year signals that Claude has cleared it — the money is run-rate spending, not experimentation.
The figure also traces the outline of Anthropic’s commercial base:
- 500 customers at $1 million or more apiece adds up to at least $500 million in annual committed spend
- A revenue base weighted toward large enterprises leaves renewals and price comparison as the primary risk variables
- Rivals now have to show enterprise-penetration evidence of matching magnitude to win over investors
That conversion — trials turning into operating cost — is the narrative propping up the valuation. It also plants a target the rest of the market can be measured against. Enterprise AI vendors lean on seat counts and pilot logos; annual spend per customer is the harder, more honest metric, and Anthropic has just made it the yardstick for any serious enterprise AI franchise.
What the Capital Will Buy
No itemized use of funds came with the announcement, but the direction is easy to infer: frontier model training, inference capacity, and expansion of the enterprise product line. February’s product cadence points the same way — Opus 4.6 arrived on February 5 with agent teams and a 1M-token context window — as demand shifts toward heavier agent workloads, and compute plus reliability are exactly the budget lines that grow in step.
Look past the “second-largest ever” headline and the round’s real confirmation is scale: enterprise demand for Claude now runs large enough that capital of this magnitude is required to support it. The question from here is not whether Anthropic has money — it is how quickly money converts into compute and product. Track capacity signals — availability, rate limits, new regions — ahead of any narrative: they expose spending earlier than any disclosure will.