Amazon is in talks to acquire Decart, an artificial intelligence startup backed by Nvidia, in a deal that could value the company at roughly $7 billion, according to a report from The Wall Street Journal. The discussions are a sign that the largest technology companies are still willing to pay up for AI businesses whose products can be plugged directly into real, revenue-generating workloads.
Decart has attracted attention for two main offerings. The first is Lucy, a model the company says can edit live video in real time — a capability that could be useful for streaming, gaming, video calls and advertising. The second is Oasis, a tool that generates simulated environments used to train and test robots and self-driving systems. Both sit in areas where demand for AI is growing quickly but where few companies have shipped products that work reliably at scale.
Why Decart is drawing interest
Live video editing is a technically demanding problem. Most AI video tools today work on clips after they are recorded, which means the user waits for processing to finish. Editing a stream as it happens requires the model to keep up with the video in real time, a challenge that has kept the technology mostly in research labs. If Decart's Lucy model works as advertised, it could open up new features across Amazon's businesses, from Prime Video and Twitch to advertising and its cloud-computing arm, AWS.
Oasis, meanwhile, fits into a broader push to train robots and autonomous vehicles in simulated worlds before letting them loose in the real one. Simulation is cheaper and safer than real-world testing, and it is one of the main bottlenecks in the race to build useful robots. Amazon already runs a large warehouse robotics operation and has invested heavily in autonomous systems, so a simulation engine could have obvious internal uses.
The reported price tag would mark a sharp step up from Decart's valuation of nearly $4 billion in May, based on PitchBook data cited by the Journal. That kind of jump in a matter of months is unusual outside of the AI sector, and it reflects how fiercely buyers are competing for a small number of startups with proven technology.
The bigger AI dealmaking picture
Amazon has been expanding its AI footprint on several fronts. The company has been revamping its Alexa voice assistant with generative AI, part of a broader effort to keep its consumer hardware relevant as AI chatbots change how people interact with devices. It has also been investing in custom chips and data centers to support AI workloads. Buying Decart would be a different kind of move — acquiring a startup with specialized models rather than building everything in-house.
The talks also highlight how complicated AI dealmaking can be. Reports of potential acquisitions often surface before terms are final, and negotiations can fall apart over price, structure or regulatory concerns. Bloomberg and other outlets have reported on the discussions, but none of the parties have confirmed a deal. Investors should treat the $7 billion figure as a reported valuation, not a completed transaction.
There is also the question of competition. Nvidia's backing of Decart gives the startup credibility and access to computing power, but it does not guarantee a sale to Amazon. Other large technology companies are hunting for similar assets, and a bidding war could push the price higher — or scare off a buyer worried about overpaying.
What it means for investors
For everyday investors, the story matters less as a single deal and more as a signal about where the AI market is heading. When a company like Amazon is willing to discuss paying a multi-billion-dollar price for a startup with a small product lineup, it suggests that demand for AI capabilities is still running ahead of supply. That supports the broader investment case for AI infrastructure — chips, data centers, cloud services — even if the specific startup changes hands or the deal never closes.
It also shows how large technology companies are using acquisitions to fill gaps rather than relying only on internal research. That can be good for shareholders if the purchases bring in talent and technology faster than building from scratch. It can be a risk if the company overpays or if the acquired technology does not integrate well. Amazon's track record on large acquisitions is mixed, and investors will want to see how any deal would fit with AWS, advertising and the devices business.
Finally, the reported valuation gap — from nearly $4 billion in May to about $7 billion now — is a reminder that private AI valuations can move quickly and are not always a reliable guide to public-market values. Investors in listed AI stocks should focus on revenue, margins and real customer adoption rather than headlines about private funding rounds.
What to watch next: whether Amazon or Decart confirm the talks, whether a rival bid emerges, and how Amazon frames the acquisition if it happens. Any deal would also face scrutiny from regulators, who have been taking a closer look at large technology acquisitions in recent years.
