Anthropic is reviewing the possibility of designing its own AI chip, Reuters reported on April 10, citing three sources.

The plan is still at an early stage, the report said. Anthropic has not settled on a specific design or formed a dedicated team and could decide to stick with its existing approach of buying chips, Reuters reported.

Anthropic currently uses Alphabet’s Google TPU and other chips to develop and run its AI software and Claude chatbot. It has also signed long-term supply agreements with Google and Broadcom. The agreements are part of Anthropic’s plan to invest $50 billion to strengthen computing infrastructure in the United States.

Demand for Anthropic’s Claude has risen sharply since the start of 2026. On that basis, Anthropic’s annualised revenue has risen from about $9 billion at the end of 2025 to more than $30 billion now, Reuters reported.

Designing its own chip is also being pursued by big tech companies such as Meta and OpenAI. Designing advanced AI chips costs about $500 million, including hiring skilled engineers and the cost of validating manufacturing processes, the sources said.