Dream Labs has raised $90 million in seed funding to develop general-purpose AI models designed to control a wide range of robots, while launching what the company describes as the first publicly available API for a robot foundation model. The funding round was led by B Capital and Kindred Ventures, with participation from A.Capital, Unusual Ventures, SV Angel, Wing VC, Scribble VC, Adrian Aoun, and other investors.
Founded in 2026 by former NVIDIA robotics researchers, Dream Labs is developing AI systems intended to give robots transferable intelligence rather than requiring each machine to be programmed for individual tasks. The company’s flagship model, Dream-1, is trained on video of people performing everyday activities and uses that information to predict what should happen next and determine corresponding robot actions.
Dream Labs said Dream-1 can control robots from different manufacturers and, without additional training, operate robots equipped with grippers. The company currently works with more than 20 hardware partners whose supported robots range in price from approximately $2,000 to $10,000.
Alongside the financing, Dream Labs introduced Dream API, which converts camera inputs and natural-language instructions into motor commands. Developers can connect compatible robots and send plain-English instructions through the API, while Dream Labs runs its models in globally distributed data centers.
The company also launched Fine-tuning as a Service, which allows developers and robot manufacturers to submit demonstration data and receive tailored models for unsupported hardware or new capabilities. Its Robot Playground lets users control a robot remotely through a browser and watch its responses live.
Dream-1 currently supports tabletop activities including picking and placing objects and folding laundry. Dream Labs plans to expand support to wheeled mobile robots and dexterous robotic hands.
The company intends to use much of the new capital for compute infrastructure as it expands the range of tasks and hardware Dream-1 can handle and increases capacity to support more users.
Dream Labs is positioning its technology around a growing gap between increasingly affordable robotic hardware and the software intelligence required to make those machines broadly useful. The company believes foundation models trained across different environments and machines could reduce the time required to deploy robots for new tasks.
KEY QUOTES:
“The hardware is ready. What’s been missing is a brain any robot can use. When you get into a rental car, you don’t relearn how to drive. You find the controls and your past driving experience does the rest. Dream-1 approaches a new robot the same way, so a new robot can do useful work in minutes, not months. And instead of edited demos, we’re giving people a working model they can test themselves, mistakes included.”
Joel Jang, Co-Founder And CEO Of Dream Labs
“A capable robot arm has gone from the price of a luxury car to the price of a high-end laptop. What it needs now is intelligence. This team did some of the defining work in robot learning at NVIDIA. Now anyone with an API key can build on it.”
Daisy Cai, General Partner At B Capital
“Our customers buy our robots to build things, and the hardest part has always been getting them to do actual work. With Dream-1, our robot was up and running in a few minutes and attempting tasks we never programmed. We can focus on building better hardware, and our customers get intelligence that keeps improving.”
Ryan Chan, Co-Founder And CEO Of MakerMods
