Anthropic has decided against acquiring AI startup Decart after conducting due diligence on a deal valued at roughly $6 billion, according to people familiar with the matter cited by Bloomberg News1,3,6.
The decision ends what would have been one of the largest acquisitions in the AI sector. Anthropic had been exploring the purchase and completed a due diligence review before stepping away, the people said4. Representatives for both Anthropic and Decart declined to comment.
The companies may still pursue other opportunities to collaborate, according to one person familiar with the matter. Bloomberg News first reported in August that Anthropic was in talks to buy Decart for around $6 billion, though nothing had been finalized and there was a possibility the talks could fall through.
Decart, an Israeli startup, develops software that can reduce the cost of training and operating AI by helping chips work more efficiently5. Reuters had previously reported that Anthropic was in talks to acquire the Nvidia Corp.-backed startup as the Claude maker explored acquisitions that could help it handle growing demand.
One report valued the potential transaction at $6 billion to $7 billion.
Anthropic rarely makes large acquisitions and has instead focused spending on computing power ahead of a highly anticipated IPO, with people familiar saying the company aims to raise as much as or more than SpaceX. The company has separately expanded into life sciences, acquiring Coefficient Bio earlier this year and launching tools such as Claude Science and the Model Hardware Standard. Anthropic also recently sealed a $35 billion cloud-computing deal with Nvidia Corp.-backed Lambda.
The IPO timeline has shifted toward mid-October, according to sources cited in a separate report.
ANALYSIS Walking away from a deal of this magnitude during active IPO preparations removes a layer of complexity from Anthropic's listing process, including valuation disputes and potential antitrust review. For Decart, the collapsed deal leaves the startup independent but without the deep-pocketed acquirer that could have accelerated distribution of its chip-efficiency software. The stated openness to future collaboration suggests the strategic rationale for combining the two companies' capabilities has not disappeared, even if the acquisition structure proved unworkable on the current timeline.