Anthropic, the private AI firm behind the Claude chatbot, is gearing up for what could be the most valuable initial public offering in history. A group of its backers told the Financial Times this week that they expect the company to go public in October with a valuation of $2 trillion or higher, easily eclipsing SpaceX's record $1.77 trillion IPO in June. That valuation would more than double the $965 billion the company was worth during its Series H funding round in May. Meanwhile, Bloomberg reported that Anthropic is in talks to acquire AI startup Decart AI for $6 billion, a move that could bolster its compute efficiency.

Anthropic confidentially filed for an IPO with the Securities and Exchange Commission in June, but has not publicly set a timeline. Rival OpenAI also filed soon after, but is not expected to go public until 2027. The awkward twist is that Anthropic is not yet profitable. To justify a $2 trillion valuation, it would need to deliver annual net income in the range of $59 billion to $79 billion, based on the average price-to-earnings multiples of the Nasdaq 100. The company is making progress: The Wall Street Journal reported that its second-quarter 2026 revenue would more than double to $10.9 billion, and it would post an operating profit for the first time. But operating profit isn't net income. It doesn't account for interest or taxes, and for a capital-intensive frontier lab, the gap could be vast.

Avery Marquez, director of investment strategies at Renaissance Capital, said reaching near operating profitability would make the huge valuation seem less outlandish. At $2 trillion, Anthropic would join an exclusive club of seven companies globally, including Nvidia, Alphabet, Apple, Microsoft, TSMC, and Amazon. Amazon, which earned $77.7 billion in net income last fiscal year, is the closest peer. Notably, Amazon's own profits are partly boosted by its investment in Anthropic, which contributed $53.4 billion in non-operating income in its most recent quarter.

Anthropic's run-rate revenue surged from $9 billion at the end of 2025 to $47 billion by mid-May. Salesforce CEO Marc Benioff estimated the run rate had reached $74.1 billion, surpassing OpenAI's $41.3 billion. Benioff praised Anthropic's enterprise hat trick: the best model (Claude), best coding agents (Claude Code), and best productivity tool (Cowork). Salesforce is an early investor and customer.

Evan Schlossberg of Neostellar Capital Corp., whose fund holds a position in OpenAI, said the key question when he sees Anthropic's S-1 will be its compute sourcing. Does Anthropic own its servers, lease them long-term, or rely on short-term leases? Owning or locking in capacity provides predictable costs and better margins. Anthropic has deals with Amazon, Google, Broadcom, and GPU access through SpaceX. The Decart acquisition would bring software to optimize chip performance and an inference team.

Marquez noted that if Anthropic goes first, it sets the benchmark for OpenAI. Anthropic's enterprise revenue is sticky and predictable, while OpenAI has massive consumer brand recognition. But both face a compute bottleneck. Schlossberg said strong demand for Anthropic's IPO would be a positive signal for OpenAI, as it suggests similar market trends. He added that no single AI model is likely to dominate global demand due to compute constraints. Marquez sees Anthropic's valuation turning up the heat on OpenAI, which will have to decide whether to double down on enterprise or focus on consumer monetization. The big hurdle for Anthropic is establishing which financial metrics matter. "The big hangup for the valuation is, what metrics make sense for this company?" she said. OpenAI will have a clear peer to compare against, but that doesn't necessarily help it.