Decart Raises $300M: Tech Leaders Back the Company as Both Customers and Investors

With funding led by Radical Ventures, Decart is building the infrastructure layer for the next generation of low-latency AI systems, through three product lines: DOS, an ultra-optimized inference and training stack that enables agents and reasoning models to run smarter and faster; and the models Lucy, its World Model for Immersive Experiences; and Oasis, its World Model for Physical AI – both powered by DOS. Today, we’re also announcing DOS 2.0, with new versions of Lucy and Oasis launching in the coming weeks.

Amazon, one of Decart’s strategic customers, anchors the partnership ecosystem. NVIDIA, Atreides Management, Valor Equity Partners,  Adobe Ventures, Toyota Ventures, and eBay Ventures join the round as new investors, alongside returning backers Sequoia Capital, Benchmark, and Zeev Ventures. Go-to-market collaborations are already underway across the ecosystem.

Some individual backers include OpenAI co-founder Andrej Karpathy, former Disney CEO Michael Eisner, the Nintendo family, and gaming investor Moritz Baier-Lentz.

‍

/A.01

MAKING WORLD MODELS A PRACTICAL REALITY

Founded by a team of optimization experts, Decart is a frontier, vertically integrated AI research lab building real-time video and world models, along with the ultra-optimized infrastructure to power them. Today, the company announced a $300 million funding round led by Radical Ventures, with participation from NVIDIA, Atreides Management, Valor Equity, Adobe Ventures, Toyota Ventures, and eBay Ventures, alongside returning investors Sequoia Capital, Benchmark, and Zeev Ventures.

The raise comes as AI moves beyond static outputs toward real-time, interactive systems that can see, respond, reason for longer and persist continuously. Decart is accelerating this transition with the models and infrastructure built to power it.

The new funding will help scale Decart’s three core product lines:

• Lucy, an immersive world model platform, with the upcoming model Lucy 2.5 focused on powering real-time experiences across gaming, e-commerce, streaming, and advertising.
• Oasis, the world’s first world model designed for Physical AI, with Oasis 3 launching soon to enable real-time world simulation.
• Decart Optimization Stack, with its new version launching today, DOS 2.0 is a next-generation inference and training platform for LLM, agentic, video, and world model workloads across all major hardware - NVIDIA GPU, Google TPU, and Amazon Trainium. DOS 2.0 delivers over 1,600 tokens per second for agentic inference (versus an industry average of around 200) and full-HD video and world model inference at up to 100 frames per second.

/A.02

ENABLED BY TECHNICAL BREAKTHROUGHS IN HARDWARE DESIGN AND INFERENCE

Decart achieves more than a 100x improvement in cost efficiency and delivers real-time, live AI performance 8x faster than any comparable system on equivalent hardware. This is enabled by the Decart Optimization Stack (DOS) - a full-stack, end-to-end platform and the first vertically integrated system designed from the ground up to make real-time AI production-deployable: spanning hardware-aware model design, kernel tooling, proprietary compilers, inference optimization, and productization. DOS runs on any hardware: NVIDIA GPUs, Amazon Trainium, Google TPUs, and integrates directly into a customer's existing infrastructure, not replacing it.

DOS is the only system capable of fully utilizing modern AI hardware, delivering up to 100 frames per second (FPS) for continuous, real-time experiences. Decart runs DOS internally to power its own models, Lucy and Oasis, making them the fastest in the world for real-time inference. The same platform is available externally, unlocking fast inference across world models, agentic AI, and LLMs.

The company is generating significant revenue through contracts with cloud providers, AI labs, and the industry's largest hyperscalers - licensing DOS to accelerate model performance and reduce compute costs at the infrastructure level. These capabilities are also foundational for robotics and autonomous vehicles, where systems must continuously interpret and react to complex, changing surroundings.
‍

/A.03

UPCOMING EXPERIENCES AND OUTLOOK  

Lucy, Decart's World Model for Immersive Experiences, is the only real-time world model running in production at scale today, powering live deployments across commerce and virtual try-on, dynamic in-video advertising, live streaming, social platforms, and gaming. Lucy transforms video and environments live, responding to user input in under 30 milliseconds. Lucy is already deployed with some of the world’s leading enterprises and is expanding continuously into new markets and industries, powered by ongoing advances in model performance, efficiency, and scale.

Oasis, Decart's World Model for Physical AI, generates interactive, physically accurate, real-time simulations, with Oasis 3 as its next release.

‍

World models are the key to moving AI from the virtual world into the physical world. Language models have taken us far, but they fundamentally operate in text - they don't understand how the physical world behaves. To move beyond that, AI needs to model environments that persist, react, and follow real-world dynamics – it needs to be able to understand structure and physics. Once systems can understand and operate within the physical world, the range of what becomes possible expands dramatically - from robotics and autonomous systems to entirely new forms of commerce and live experience. The most immediate test is what happens when you put it in front of a game designer, an advertiser, or an engineer building a robot - and watch what they reach for.

– Dean Leitersdorf, CEO & co-founder, Decart

/A.04

VOICES FROM OUR INVESTORS AND PARTNERS  

Decart’s world model infrastructure operates at the intersection of frontier model development and next-generation, AI-optimized systems powered by NVIDIA’s accelerated computing platform. NVIDIA’s investment supports Decart’s vertically integrated approach across hardware, model architecture, and productization, enabling the development of AI systems that operate at a physical-world scale.

Building on DOS, Decart formed a partnership with Amazon Web Services spanning a commercial agreement and joint go-to-market initiative to bring real-time world models to AWS customers at scale. Through deep collaboration with AWS teams, including Annapurna Labs, Decart has become one of the first companies to deploy real-time AI models of this class and scale on Trainium, Amazon’s custom AI chip, with Lucy2 now running on Trainium3. Decart adapted its model architecture to take full advantage of Trainium’s chip design- a co-design approach that makes compute-intensive, real-time applications practical at scale. As a member of the Amazon Partner Network, Decart's API gives the AWS enterprise customer base a direct on-ramp to build with real-time, interactive AI across media, commerce, advertising, and physical AI.

“Decart is the kind of partner we love working with: technologically fearless, fast-moving, and deeply innovative. They were among the first in the world to optimize on Trainium3 and have embraced our newest software libraries. The results speak for themselves. Their latest model, Lucy2, exceeds 80% Model FLOPS Utilization, meaning more of the chip's raw power is doing real, productive work. In an industry where most workloads waste a significant share of the hardware they run on, that's real money saved and real time back. It's exactly the kind of efficiency Trainium was designed to unlock“

– Nafea Bshara, VP, Amazon’s Annapurna Labs

The breadth of Decart's investor base reflects the range of industries its technology could fundamentally transform. NVIDIA, Adobe Ventures, eBay Ventures, Toyota Ventures, Disney's former CEO, and the Nintendo family each bring a different lens to the table. Decart is already live with partners and customers across consumer electronics, creative tools, entertainment, gaming, industrial systems, and physical environments. As a strategic customer, Amazon has a deep technical collaboration with Decart on AWS Trainium. Radical Ventures, Sequoia Capital, Benchmark, Atreides Management, and Valor Equity Partners round out the institutional base.

Decart is the engine powering the future of real-time media and physical AI. Decart makes State of the Art real-time video economically viable at massive scale. While the AI industry has considered this technological achievement years away, Decart has built a software platform that combines proprietary chip optimization, SOTA world models, and customer API that has closed the gap between research and self-serve product platform in a way no one else has. While the applications across media and robotics are endless, the thesis is simple: Decart has built the foundational layer that makes this entire category possible at scale.
‍– Jordan Jacobs, Managing Partner, Radical Ventures

The scale and diversity of Decart’s investor coalition signal something larger: Decart is emerging as general-purpose AI infrastructure, built to serve entire industries rather than any single market. This funding will be used to scale Decart's infrastructure platform, expand enterprise distribution through strategic partnerships globally, and accelerate development of its real-time world models.

‍

About Decart

Decart is a fully vertically integrated frontier AI research lab, founded in 2023 and building state-of-the-art, real-time world models and the ultra-optimized infrastructure to power them. At the core of the company is the Decart Optimization Stack (DOS), spanning hardware-aware model design, data pipelines, inference optimization, and productization, enabling real-time video generation at a fraction of industry cost. Backed by Radical Ventures, NVIDIA, Sequoia Capital, Benchmark, Zeev Ventures and many of the industry’s leading strategic and angel investors, Decart has raised over $450 million to date.