Amazon Reportedly in Talks to Buy Decart for About $7B
Amazon is reportedly discussing a roughly $7B acquisition of Decart. The talks are not signed, following earlier ended negotiations with Anthropic.
The Wall Street Journal reports that Amazon is in talks to acquire AI infrastructure startup Decart for about $7 billion. No agreement has been announced, and the discussions could still change or end without a transaction.
Status: reported acquisition talks — not a signed or completed deal.
Deal snapshot
| Item | Detail |
|---|---|
| Potential buyer | Amazon |
| Target | Decart |
| Reported value | About $7 billion |
| Status | Negotiations reported by The Wall Street Journal |
| Last disclosed financing | $300 million in May 2026 |
| Funding to date | More than $450 million, according to Decart |
| Last reported valuation | Nearly $4 billion in May 2026 |
Why Decart could matter to Amazon
Decart is best known publicly for its real-time world models, including Lucy and Oasis, but its strategic value may sit deeper in the stack. The company says its Decart Optimization Stack helps AI developers run demanding training and inference workloads across different accelerators, including Nvidia GPUs, Google TPUs, AMD chips and Amazon's Trainium.
That portability addresses a structural problem in AI infrastructure: models are expensive to tune for one hardware architecture, while access to leading chips remains constrained. A buyer that owns both cloud infrastructure and custom silicon could use Decart's software to reduce the switching cost for customers considering alternatives to Nvidia.
The relationship is not hypothetical. Decart offers single-tenant deployments on AWS Trainium and Nvidia Blackwell, and its own product materials say the Lucy video model can generate frames at sub-50-millisecond latency on Trainium.
A second reported sale process
The Amazon report follows earlier coverage linking Decart to Anthropic. Israeli reporting said Anthropic had discussed a transaction valued around $6 billion to $7 billion, largely in shares, but a later report said those talks ended. Neither Decart nor Anthropic announced a signed deal.
That history makes precision important: today's development is evidence of continuing strategic interest, not proof that Amazon has won an auction. Price, consideration mix, employee retention terms and regulatory conditions remain unknown.
The valuation logic
Decart announced a $300 million round in May led by Radical Ventures, with Nvidia, Adobe Ventures, Toyota Ventures, eBay Ventures and others participating. The company says that financing took total funding above $450 million. Contemporary reporting put its valuation near $4 billion.
A $7 billion purchase would therefore represent a roughly 75% premium to that reported private-market mark. The premium would be easier to justify if Amazon values Decart as both:
- an optimization layer for AWS's custom AI chips; and
- a research team building low-latency video and physical-AI models.
The proposed price would be harder to defend if the value rests mainly on early world-model demonstrations rather than durable enterprise workloads.
What investors should watch
- Confirmation: a company statement, regulatory filing or signed agreement.
- Consideration: cash versus Amazon shares and any performance-based component.
- AWS integration: whether Decart remains hardware-neutral or becomes primarily a Trainium optimization layer.
- Customer concentration: how much of Decart's current commercial activity already depends on Amazon.
- Regulatory review: a large AI infrastructure acquisition would attract scrutiny over cloud and compute concentration.
For now, the correct description is reported talks for a possible $7 billion acquisition. It is not a completed exit.
Sources
Follow Venture Capital Tracker in Google
Add VCT as a preferred source to make our venture-capital coverage easier to find in Google Search.
Last updated:
Editorial note: AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.
