Dream Labs raises $90 million to bring robot control models to developers
The AMW Read
The $90 million seed round and accessible robot-control API meaningfully update the embodied-model player map, with segment-level implications contingent on reliable transfer across hardware.
Dream Labs raises $90 million to bring robot control models to developers
Dream Labs raised $90 million in seed funding led by B Capital and Kindred Ventures and launched Dream API, which converts camera inputs and natural-language instructions into robot motor commands. Founded in 2026 by former NVIDIA robotics researchers, the company trains its Dream-1 model on video of people performing everyday activities. Dream Labs says it works with more than 20 hardware partners, with supported robots priced at roughly $2,000 to $10,000. Current tasks include tabletop picking, placing and laundry folding.
The company is pursuing a horizontal software layer for physical AI: intelligence that transfers across robot manufacturers rather than requiring separate programming for every machine and task. Dream Labs says Dream-1 can operate robots equipped with grippers without additional training. Its cloud-hosted API and Fine-tuning as a Service create two routes to adoption: direct access for compatible hardware and customization using demonstration data. That positions the business around deployment effort and hardware coverage, although the reported capabilities do not establish reliability across varied operating environments.
For builders, the concrete opportunity is to test whether shared robot intelligence reduces integration work on affordable hardware. Dream Labs offers a browser-based Robot Playground for live remote testing; deployment evaluation should also measure task completion, intervention frequency and response latency on the intended robot. For investors, the key question is whether broader hardware support can become a durable advantage while the company funds model serving and expansion. Dream Labs says much of the new capital will go toward compute infrastructure, making operating costs relevant alongside capability gains.
