Anthropic, the developer of the Claude AI model, expects its initial public offering to match or beat SpaceX’s record-setting IPO size of US$75 billion, which grew to US$86.2 billion with overallotments [1, 2, 3, 4]. The company plans to file IPO documents publicly as soon as late August 2026 [1, 2, 3, 4].
In May 2026, Anthropic raised US$65 billion in funding at a US$965 billion valuation, surpassing OpenAI’s US$852 billion valuation from March 2026 [1, 2, 4]. Its preliminary Q2 2026 revenue exceeded US$11.5 billion, a sharp increase from US$787 million in Q2 2025 [1, 2, 4]. By the end of July 2026, Anthropic’s revenue run rate had reached US$65 billion [1, 2, 4].
Despite positive adjusted operating income in Q2 2026, the company reported a net loss of nearly US$42 billion in 2025, about five times the previous year’s loss [1, 2, 4]. The high costs of training advanced AI models require significant computing resources. Anthropic has a multi-billion-dollar deal with SpaceX to supply computing resources over the next several years [1, 2].
The IPO underwriting group is led by Morgan Stanley, Goldman Sachs, and JPMorgan Chase [4]. Anthropic is considering a dual-class share structure with super voting rights, enabling CEO Dario Amodei and a co-founder to maintain controlling decision power after the IPO [4].
Recent investor briefings, led by Chief Financial Officer Krishna Rao, skirted questions regarding the company’s exact valuation ahead of the IPO [1]. The upcoming public filing will provide more details about Anthropic’s financials and ownership structure.
