#anthropic滚动12个月营收增长约1200%
【Revenue up 1,200% in a year—Anthropic really has turned AI into a money-printing machine】
One of the most staggering figures in Anthropic’s IPO filing with the U.S. SEC is that revenue for the 12 months through September 2026 grew about 1,200% year over year, while its headcount grew only about 170%. Revenue per employee rose from $1.4 million to more than $6.6 million. Products like Claude Code are turning demand for AI into real revenue, and businesses are willing to keep paying—far more convincing than simply racking up users.
But soaring revenue doesn’t mean profits will take off too. In the first half of 2026, $ANTHROPIC revenue reached $16.3 billion, while spending on computing power and infrastructure also hit about $8.9 billion. The AI business has proven it can be commercialized, but the more people use these models, the more staggering the computing bills become.
The questions now are whether revenue can keep growing, whether computing costs can come down, and whether customers will cut spending in response to competition and lower prices. Growth of 1,200% is stunning, but it can’t last forever. If revenue growth starts to slow and costs don’t come down, this money-printing machine will face a valuation test. For now, this is stronger evidence that commercial demand for AI is booming—not proof that high valuations will always be justified.
【Revenue up 1,200% in a year—Anthropic really has turned AI into a money-printing machine】
One of the most staggering figures in Anthropic’s IPO filing with the U.S. SEC is that revenue for the 12 months through September 2026 grew about 1,200% year over year, while its headcount grew only about 170%. Revenue per employee rose from $1.4 million to more than $6.6 million. Products like Claude Code are turning demand for AI into real revenue, and businesses are willing to keep paying—far more convincing than simply racking up users.
But soaring revenue doesn’t mean profits will take off too. In the first half of 2026, $ANTHROPIC revenue reached $16.3 billion, while spending on computing power and infrastructure also hit about $8.9 billion. The AI business has proven it can be commercialized, but the more people use these models, the more staggering the computing bills become.
The questions now are whether revenue can keep growing, whether computing costs can come down, and whether customers will cut spending in response to competition and lower prices. Growth of 1,200% is stunning, but it can’t last forever. If revenue growth starts to slow and costs don’t come down, this money-printing machine will face a valuation test. For now, this is stronger evidence that commercial demand for AI is booming—not proof that high valuations will always be justified.