Amazon Decart Acquisition Talks Hit $7B in Chip Race

Amazon is in advanced talks to acquire Decart, an Israeli-founded artificial intelligence startup, in a deal valued at around $7 billion, according to reporting from the Wall Street Journal via Reuters on October 11, 2026. If completed, the Amazon Decart acquisition would rank among the largest AI deals of the year and mark another major consolidation move in the industry's escalating compute arms race. Both companies declined to comment on the reports.

Decart was founded in 2023 by three engineers and has moved at breakneck speed from research lab to one of the most sought-after startups in AI infrastructure. The company was valued at nearly $4 billion as recently as May 2026, according to PitchBook data, and raised $300 million led by Radical Ventures, with Nvidia joining as a new investor. That means Amazon would be paying a hefty premium — a signal of just how badly Big Tech wants control of the layers between raw chips and the AI models that run on them.

What Decart actually does

Decart is not a chipmaker in the traditional sense. Its flagship product, the Lucy model, edits live video in real time — a technology that attracted Nvidia's backing and puts Decart at the bleeding edge of generative video. The company's Oasis model takes a different approach: it generates fully simulated 3D environments used to train and test robotics and autonomous-driving systems, letting engineers run millions of virtual miles without ever touching a real road.

But the part of Decart that Amazon likely covets most is its software for chip portability: tools that let developers move AI workloads across chips from different vendors with minimal friction. That matters because the AI industry currently runs almost entirely on Nvidia hardware. Companies large and small are desperate for ways to diversify their chip supply and cut costs, and portable software is the key that unlocks competition in the silicon market.

Why Amazon wants in now

The connection between the two companies is already deep. In 2025, Amazon gave Decart early access to its Trainium3 custom AI chip, a sign that the relationship had been warming long before acquisition talks began. Owning Decart outright would let Amazon pair its custom silicon with software that makes it easy for developers to adopt — attacking Nvidia's dominance from both the hardware and the software side at once.

Amazon is not the only suitor to have circled Decart. In August 2026, Anthropic explored a roughly $6 billion acquisition but walked away after due diligence, according to reports. The fact that talks with Amazon have advanced further suggests the cloud giant sees something worth paying more for — likely the combination of Decart's talent, its real-time video tech, and the strategic value of chip portability in Amazon Web Services' battle for AI cloud market share. For AWS, which has invested billions building its own Trainium and Inferentia chip lines, software that smooths the path for customers to migrate off Nvidia hardware could become a powerful differentiator in winning enterprise AI workloads.

Why this matters to you

Here is the bigger picture: the AI compute arms race is consolidating fast. Whoever controls both the chips and the software layer on top of them controls the economics of AI itself — and that shapes everything from the price of the tools you use daily to who gets to build the next generation of apps. A $7 billion Amazon Decart acquisition would hand one of the world's biggest cloud providers a full stack play: custom chips, the software to run anything on them, and the video and simulation tech to power entirely new experiences.

Whether the deal closes remains to be seen. People familiar with the matter say an agreement could come soon if the current talks hold, but as the Anthropic episode showed, billion-dollar AI deals can collapse at the last minute. For now, further reporting on the talks and additional coverage of Decart's chip portability technology suggest this is one acquisition saga worth watching closely — because the winner of the chip wars may decide the price of your AI future.