Dream Labs raises $90M and launches Dream API
Dream Labs has closed a $90 million seed round and released Dream API, a platform that lets robots learn from video and execute tasks using plain‑English instructions. The AI model, Dream‑1, works out‑of‑the‑box on grippers and mobile robots from multiple vendors.
Radwan AbdulhadiEditor and technology writer
Funding & Launch
Dream Labs announced a $90 million seed financing on October 22, 2026, according to The Robot Report. The round was led by B Capital and Kindred Ventures, with participation from A.Capital, Unusual Ventures, SV Angel, Wing VC, Scribble VC and Adrian Aoun, among others. The company also unveiled Dream API, a publicly available interface that converts a robot’s camera feed and natural‑language instructions into motor commands. Interested developers can read more about the launch at The Robot Report.
Cross‑Manufacturer AI Model
The flagship Dream‑1 model works across robots from different manufacturers, the startup says, enabling zero‑shot learning and out‑of‑the‑box control of grippers and other hardware. Trained on video of people performing everyday tasks, Dream‑1 can predict next actions and transfer its capabilities to new hardware without any data. ‘The hardware is ready. What’s been missing is a brain any robot can use,’ said Joel Jang, co‑founder and CEO of Dream Labs. Developers can connect any robot with grippers and watch it execute tasks, leveraging the model’s cross‑platform intelligence.
Developer Access & API
Developers can sign up today and start using Dream Labs’ AI products. The Dream API translates visual inputs and natural‑language prompts directly into robot actions, allowing a robot to carry out a task as soon as it is connected. For unsupported hardware or new capabilities, developers and robot makers can submit a small batch of demonstration data to receive a tailored model within hours. Dream Labs also offers a public robot playground where anyone with a browser can type commands to a robot in its San Francisco office and watch it respond live, no purchase required. This includes wheeled mobile robots and dexterous hands.
Real‑World Impact & Partnerships
Dream Labs currently supports robots priced between about $2,000 and $10,000, with Dream‑1 needing no extra data to control a two‑armed robot with parallel grippers. The startup works with more than 20 hardware partners, refining the model for each robot to ensure optimal performance. ‘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,’ said Daisy Cai, general partner at B Capital. Early users like MakerMods, which sells $2,000 bimanual robot arms, reported that their robot was up and running in minutes with Dream‑1, attempting tasks never programmed. The team previously did defining work in robot learning at NVIDIA, and now anyone with an API key can build on that foundation.
Future Roadmap
The company plans to use the new funding primarily for compute as it expands the tasks and robots Dream‑1 can handle and scales to support more users. ‘Our customers buy our robots to build things, and the hardest part has always been getting them to do actual work,’ said Ryan Chan, co‑founder and CEO of MakerMods. With Dream‑1, the robot was up and running in a few minutes and attempting tasks we never programmed. This allows the hardware focus to improve while customers receive intelligence that keeps improving. For more coverage of funding and investments in robotics, read more at The Robot Report’s investments section.
Sources
- The Robot ReportDream Labs raises $90M and launches Dream API
Continue chronologically
Related entity coverage
Related stories
Meta Reportedly Builds Humanoid Robot Prototype for Chores
Meta has reportedly built a humanoid robot prototype codenamed Max designed for household chores, with work already underway on a second version.
Teradyne Robotics and Sereact Partner for Physical AI
Teradyne Robotics has partnered with Sereact to scale physical AI solutions across retail and logistics, building on hundreds of live deployments.
Jensen Huang Unveils Plug-In Data Centers for Europe at GTC
At Nvidia's GTC in Berlin, Jensen Huang unveiled the modular DSX Pod data center and new humanoid robotics initiatives to target Europe's AI infrastructure lag.
SimReady Central Launches With Over 2,000 OpenUSD Assets
Discover SimReady Central Library on Hugging Face, offering over 2,000 free OpenUSD assets for robotics simulations and AI factory digital twins.
