Anthropic Shifts from $7B Acquisition of MatX to Design Partnership Exploration

Anthropic, a prominent AI company, has walked away from its proposed $7 billion acquisition of MatX, an AI chip startup founded by former Google engineers. Instead, the two companies are reportedly looking into a design partnership, reflecting a significant change in Anthropic's hardware strategy.

The initial acquisition bid was notable, as it represented a considerable premium over MatX's current valuation, which surged to multi-billion dollars following a $500 million Series B funding round in February 2026. This funding round was led by Jane Street and Situational Awareness LP, establishing MatX as a key player in the AI chip market.

Anthropic's interest in MatX stemmed from its need for specialized hardware to efficiently run its Claude family of models. Acquiring MatX would have provided Anthropic access to a team of experienced chip designers and the MatX One chip, engineered for large language model workloads. However, the reasons behind Anthropic's decision to abandon the acquisition remain unclear, as neither company has publicly commented on the matter.

In a broader context, Anthropic is assembling its own internal custom silicon team and investing up to $5 billion in partnerships to diversify its chip supply beyond Nvidia. This indicates a long-term commitment to developing its hardware capabilities. The potential design partnership with MatX could take various forms, including licensing chip architecture or co-developing processors optimized for Anthropic's models.

For MatX, remaining independent may prove beneficial, allowing it to partner with Anthropic while also pursuing sales to other AI labs and enterprises. The willingness of Anthropic to offer $7 billion underscores the increasing value AI labs place on custom silicon capabilities.