According to the latest news from Reuters, generative AI giant Anthropic is evaluating the feasibility of developing its own AI chips. With the surge in demand for its Claude model in 2026, the company aims to enhance its control over computing power by developing hardware in-house, reducing reliance on external suppliers.

Revenue Surge Drives Strategic Transformation

Thanks to the strong performance of the Claude series models in global markets, Anthropic's financial data has seen a significant growth in the past year. As of April 2026, its annualized revenue has officially exceeded $30 billion, more than tripling from $9 billion at the end of 2025. This solid financial foundation provides confidence for its entry into the high-barrier chip development field.

Although the initial cost of developing its own chips is expected to reach $500 million and involves complex engineering challenges, Anthropic's move is aimed at competing with OpenAI and Meta. The company has already established long-term partnerships with Google and Broadcom, deepening its existing TPU layout while gradually advancing its $50 billion computing infrastructure investment plan.

Addressing Global Chip Supply Shortages

The self-developed chip plan is still in the evaluation stage, without a dedicated team formed yet. The company currently still heavily relies on general-purpose chips and customized computing power provided by Amazon and Google. However, as the scale of AI model training continues to expand, fluctuations in external chip supply have become a key factor limiting the iteration speed of its large models.

Industry experts analyze that self-developed chips can not only significantly reduce long-term operating costs but also enable fundamental optimization for its proprietary algorithms, maximizing computing efficiency. Facing the unresolved "chip shortage" issue globally, Anthropic is trying to gain more leverage in top AI competitions through the deep integration of software and hardware.