Business•

Amazon Closes In on a Deal Worth Up to $7.5 Billion for Israeli AI Firm Decart, Nearly Double Its May Valuation

Amazon is close to buying Israeli AI startup Decart for up to about $7.5 billion, a deal aimed at strengthening its in-house AI chips against Nvidia.

Written By :MINRK
Amazon Closes In on a Deal Worth Up to $7.5 Billion for Israeli AI Firm Decart, Nearly Double Its May Valuation

Photo by Zach M on Unsplash

1. Amazon Decart Acquisition Nears Signing at Up to $7.5 Billion

Amazon is close to acquiring Decart, an Israeli artificial intelligence startup, in a deal valued at up to about $7.5 billion, according to people familiar with the negotiations, as of November 21, 2026. The Amazon Decart acquisition, which neither company has announced, could be finalized within days and would rank among the most expensive purchases of a young AI company to date.

The price would represent a sharp step up from the valuation Decart carried only months ago. It also signals how much Amazon is willing to pay for technology that could help its own AI chips compete more effectively with those made by Nvidia, the company that dominates the market for AI computing hardware.

Because the deal has not been officially announced, its final terms could still change. Talks of this size often shift in their last stages as the parties settle details on price, payment structure and the obligations placed on key staff.

The negotiations come at a time when the largest technology companies are spending heavily to secure an edge in artificial intelligence. Much of that spending goes to the physical infrastructure behind AI services: data centers, networking equipment and, above all, the specialized processors needed to train and run large models. Control over that hardware stack has become a central competitive question for cloud providers.

2. How the Deal Value Climbed

The figure under discussion has risen over recent weeks. In October, the talks between Amazon and Decart were described as centering on a price of about $7 billion. The latest terms push the potential total to as much as $7.5 billion.

The payment would combine cash, Amazon stock and deferred payments. That kind of structure is common in large technology acquisitions. Spreading part of the consideration over time allows a buyer to keep the founders and core engineers motivated after the sale closes, while letting the headline value reflect the full amount that could eventually be paid.

Under the terms, Decart's founders have committed to remaining with the business for five years, with a significant portion of their payment due at the end of that period. For Amazon, keeping the people who built the technology is likely as important as the technology itself, since the company is paying largely for expertise that is difficult to replicate.

The use of Amazon stock as part of the payment also ties the sellers' eventual return to Amazon's own performance. If the acquisition helps strengthen Amazon's AI business, the value of that stock component could rise over the years in which the founders remain committed to the company, aligning their interests with those of their new employer.

3. What Does Decart AI Do?

Decart AI builds real-time world models and optimization software that helps AI models train and run efficiently across different chip architectures, including Amazon's. In simple terms, the company works on two fronts: generating interactive simulated environments and video on the fly, and making AI workloads run better on a variety of hardware.

World models are AI systems that learn how environments look and behave, so they can generate scenes that respond to a user's actions in real time. The technology has drawn attention for uses in gaming, simulation and interactive video, and it is considered one of the more demanding areas of AI in terms of computing power.

For Amazon, the optimization work is likely the more strategic asset. Most AI software today is written with Nvidia's chips and software ecosystem in mind, which makes it costly and complicated for customers to move their workloads elsewhere. Tools that make models run efficiently on alternative chips lower that barrier.

Decart has run its models on Trainium, the custom AI training chips developed by Amazon Web Services (AWS), Amazon's cloud computing division. Its work on that hardware gives Amazon a working example of how its chips can handle demanding, real-time AI applications.

Training and running AI models are distinct tasks. Training is the process of teaching a model by feeding it vast amounts of data, which requires enormous computing power over weeks or months. Running a model, often called inference, is the work of answering requests once training is complete. Software that improves efficiency at either stage can translate directly into lower costs for the companies operating the hardware and the customers renting it.

4. An Existing Customer Relationship With Amazon

The two companies are not strangers. Amazon was already a strategic customer of Decart before acquisition talks began, which means Amazon's engineers have had direct experience with the startup's software and how it performs on AWS hardware.

That history reduces some of the uncertainty that usually accompanies an acquisition. Rather than buying a technology it has only evaluated from a distance, Amazon would be taking full ownership of a partner whose products it has already put to use. It also helps explain why Amazon may be willing to pay a premium: it has seen the technology work in its own environment.

5. From a $4 Billion Round to a Potential $7.5 Billion Exit

Decart's valuation has moved quickly. In May 2026, the company raised $300 million in a funding round that valued it at roughly $4 billion. Across its funding history, the startup has raised more than $400 million in total.

If the $7.5 billion figure holds, Decart's value would nearly double in about six months. That pace would be unusual even by the standards of the current AI investment cycle, in which valuations for promising startups have risen rapidly as large technology companies compete for talent and intellectual property.

For Decart's investors, a sale at that price would deliver a substantial return in a short period. For the broader AI startup market, it would add another data point showing that the largest cloud and technology companies are prepared to pay heavily for specialized capabilities that strengthen their position in AI infrastructure.

The speed of that revaluation also reflects the specific nature of Decart's work. Companies that can help make AI hardware more efficient sit at the center of the industry's most expensive problem, the cost of computing power, which makes their technology attractive to chipmakers, model developers and cloud providers alike.

6. Nvidia and Anthropic Talked to Decart First

Amazon was not the first company to explore a deal. Decart previously held discussions with Nvidia, and it also negotiated a possible acquisition with Anthropic, the developer of the Claude AI models, in talks that valued the deal at around $6 billion. Those discussions with Anthropic ended in September 2026.

The interest from both companies underlines how valuable Decart's technology is seen to be across the AI industry. Nvidia, as the leading supplier of AI chips, has a clear interest in tools that shape how models run on hardware. Anthropic, as a major developer of AI models, depends on access to large amounts of efficient computing power.

For Amazon, owning Decart outright would also have a defensive benefit. Bringing the company in-house would prevent rivals from using its technology to improve their own AI systems and hardware platforms, keeping its capabilities focused on Amazon's chips and cloud services.

7. Why Amazon Wants Decart Now

The acquisition fits a long-running effort by AWS to reduce its dependence on Nvidia. Like other major cloud providers, Amazon buys large volumes of Nvidia processors to power AI services for its customers, and it has spent years developing its own chips, including Trainium, as an alternative.

Custom chips can lower costs and give a cloud provider more control over supply, pricing and performance. Their success, however, depends on whether developers can easily run their models on them. Hardware alone does not win customers if the software tools around it lag behind what Nvidia offers.

Amazon is not alone in pursuing this strategy. Other large cloud providers have also invested in designing their own AI processors in an effort to manage costs and reduce reliance on a single supplier. In that contest, the quality of the surrounding software can matter as much as the raw performance of the chips, since customers tend to choose platforms where their existing models run with the least friction.

That is the gap Decart's optimization technology is meant to help close. By making it easier to train and run AI models efficiently on Trainium, the startup's tools could make Amazon's chips more attractive to customers who currently default to Nvidia hardware, which in turn would strengthen AWS's position in the race to supply AI computing capacity.

The deal also reflects the scale of Amazon's broader AI ambitions. A purchase worth several billion dollars is large by the standards of startup acquisitions, but it is one piece of a much wider investment program aimed at securing Amazon's place in AI infrastructure.

8. What Comes Next for the Amazon Decart Deal

The next step is a signed agreement and a formal announcement from the companies, which could come within days if the negotiations stay on track. Until then, the size and structure of the deal remain subject to change.

For Decart, the next phase would mark a shift from independent startup to part of one of the world's largest technology companies. The five-year commitment from its founders suggests that the business is expected to continue under the leadership of the people who built it, at least for the duration of that period.

Investors and industry observers will be watching for details on how Amazon plans to fold Decart's teams and products into AWS, and whether the startup's world models and optimization tools will be offered to cloud customers or used mainly to improve Amazon's own chip platform. Those decisions will show how central the acquisition is to Amazon's plan to challenge Nvidia's grip on AI hardware.

Related Articles

NEWSLETTERS

Don't miss another story.

Subscribe to the MINRK Newsletter today.

Tick the newsletters you want. No selection? You'll get MINRK Headlines.

By signing up, you will receive emails about MINRK products and you agree to our terms of use and privacy policy.

MINRK Daybook

Market analysis for crypto traders and investors.

EVERY WEEKDAY

MINRK for Advisors

Defining crypto, digital assets and the future of finance for financial advisors.

EVERY THURSDAY

MINRK Protocol

Exploring the tech behind crypto one block at a time.

WEEKLY

MINRK Long & Short

A must read for institutions. Insights, news and analysis delivered weekly.

EVERY WEDNESDAY

MINRK Headlines

The biggest crypto news and ideas of the day.

EVERY WEEKDAY

MINRK Policy

Examining the intersection of cryptocurrency and government.

WEEKLY

MINRK Research

Join thousands of readers who rely on MINRK for data-driven insights on the latest digital asset trends.

MONTHLY