Anthropic lost more than $80bn (£59.8bn) in the first half of the year as the artificial intelligence (AI) giant burns through cash before its blockbuster stock market float.
The company behind the Claude chatbot reportedly posted a net loss of $80.7bn in the first half of 2026, a substantial increase from the $13bn it lost in the same period last year.
Anthropic said the increased losses largely reflect accounting adjustments associated with its financing agreements.
However, the tech giant has faced a dramatic increase in costs as it races to develop its AI models and get ahead of rivals such as ChatGPT-maker OpenAI.
Anthropic spent nearly $9bn on computing capacity and infrastructure in the first half of the year – more than it spent in the whole of 2025.
The company has entered into cloud computing contracts totalling approximately $450bn over the next seven to 10 years. Its overall costs could hit $750bn as it continues to negotiate deals.
These agreements include $111bn with Google Cloud, $110bn with Amazon Web Services, $31bn with Microsoft and up to $86bn to rent Nvidia chips from Nscale and Lambda.
Anthropic has also entered a contract exceeding $160bn to lease Google chips via Broadcom.
The figures, contained in a draft prospectus obtained by the Wall Street Journal, come as Anthropic gears up for an initial public offering next month that could value it at as much as $2tn.
The company generated approximately $16.3bn in revenue in the first six months of the year.
The figure marks a near-15-fold jump from $1.1bn during the same period in 2025 and eclipses its total revenues of $4.6bn last year.