Anthropic은 11월 기업공개를 계획 중이며, 최대 $100 billion의 밸류에이션을 목표로 합니다.

Anthropic is in discussions about an initial public offering that could raise up to $100 billion at approximately a $2 trillion valuation, with Nvidia considering an investment of as much as $10 billion. The IPO terms remain under discussion and could shift, but the company is not expected to complete the offering before the U.S. midterm elections in November.

A $100 billion stock sale would exceed SpaceX's June IPO proceeds of roughly $86.25 billion by about 16 percent. However, this capital raise masks a more pressing challenge: Anthropic has committed to roughly $518 billion in AI infrastructure spending over the next decade.

These commitments break down into two major categories. Spending with hyperscalers Alphabet, Amazon, and Microsoft totals about $252.5 billion and includes minimum requirements even if Anthropic uses less capacity. Broadcom-linked equipment leases account for another $161.2 billion, with limited cancellation options. Together, these total approximately $413.7 billion—meaning a $100 billion IPO would cover only about 24 percent of this largely fixed portion of planned compute spending.

Anthropic's operating fundamentals are not dire. The company reported a net loss of approximately $42 billion in 2025, but roughly $34 billion of that stemmed from accounting charges tied to financing instruments. Operating losses were significantly lower at about $8 billion, and the company achieved positive adjusted operating income in the second quarter of 2026 based on preliminary figures.

The deeper concern is the inflexibility between revenue and costs. Usage-based spending accounted for 82.6 percent of 2025 revenue, with just two clients representing 24 percent of total revenue. Most large customers lack long-term commitments and can reduce Claude usage relatively quickly. Meanwhile, approximately 80 percent of Anthropic's $518 billion infrastructure plan consists of effectively fixed commitments—either non-cancelable or requiring payment regardless of usage—leaving the company with far less flexibility to adjust its future compute expenses in response to demand changes.