Dream Labs raises $90M seed round and releases Dream-1 world action model

Dream Labs announced the completion of a $90 million seed round funding and the release of the world action model Dream-1. This model can control previously unseen robots and perform new tasks, with an operational cost as low as $3 per hour. The model achieves cross-robot generalization by predicting the end-effector pose, although it has limitations in control accuracy and task adaptability. The Dream API is now open to global users, supporting simulation testing and fine-tuning services.

Dream Labs, founded by former NVIDIA DreamZero team members, has released the embodied model Dream-1, which supports zero-shot task execution across different robots. The model demonstrates strong generalization capabilities in both simulated and real-world environments, achieving the top score in the RoboDojo-RC Tier 1 test. Dream-1 learns to predict visual scenes and actions through a world action model (WAM), allowing it to perform tasks on previously unseen dual-arm robots without additional adaptation.

Dream Labs, founded by the NVIDIA DreamZero team, has launched Dream API, the first publicly available robot brain API. Dream-1, the model behind the API, enables zero-shot control of robots using natural language commands, with support for over 20 hardware partners. It outperforms several large models in benchmark tests, though it is still in beta with known limitations.

Dream Labs has released Dream-1, a world action model that enables robots to perform tasks without prior training or customization. The model can control robots it has never seen, using a shared interface for end-effector control. Dream-1 is available via an API and offers cost-effective deployment, with performance benchmarked against existing models like GPT-6 Astra and Gemini 4 Argon.

Dream Labs, founded by the NVIDIA DreamZero team, has launched Dream API, the first publicly accessible robot brain API. Dream-1, the model behind the API, enables zero-shot control of robots using natural language commands, with support for over 20 hardware partners. It outperforms many large language models in benchmark tests, with lower cost and faster execution. The model is still in research beta, with known limitations in handling complex kinematic constraints and novel robot designs.