1X launches world model enabling NEO robot to learn tasks by watching videos - The Robot Report
The Palo Alto, Calif.-based company said it has designed NEO for household use. The humanoid is available through 1X’s early access program for $20,000, which includes priority delivery in 2026. A subscription model will also be available for $499/month.
Key takeaways
The most recent developments in humanoid robotics show a rapid expansion of both commercial deployments and research breakthroughs. Agibot, the 2023‑founded company that shipped more than 5,100 units in 2025 and captured a 39 percent global market share, announced the launch of its embodied‑intelligence platform in Malaysia, marking the first of several Asia‑Pacific rollouts for 2026. In the United States, Fauna Robotics introduced “Sprout,” a 1‑meter‑tall, soft‑foam‑cushioned humanoid designed for home interaction and already being hand‑delivered to early customers such as Disney and Boston Dynamics, while Oversonic Robotics brought its cognitive humanoid “RoBee” to the U.S. market for healthcare and advanced‑manufacturing use. Hyundai, together with Boston Dynamics, confirmed plans to field thousands of Atlas AI‑powered humanoids at its Georgia EV plant beginning in 2028, though the company’s labor union warned that deployment will require collective‑bargaining agreements. At the Davos robotics forum, experts highlighted that the next hurdle for humanoids is robust perception and manipulation, noting the need for robots that can learn directly from human coworkers rather than rely on costly demos. OpenAI has quietly rebuilt a robotics lab in San Francisco, employing about 100 data collectors to teach a humanoid platform household tasks via tele‑operated Franka arms, and is preparing a second site in Richmond, California. Tesla’s Optimus humanoid is slated to move from factory prototypes to consumer sales by the end of 2026, with Elon Musk promising high reliability and applications in elder‑care and domestic assistance. Finally, 1X Technologies released a new “World Model” that lets its NEO humanoid learn tasks by watching videos, offering a subscription‑based service for household use, while Tesollo unveiled a lighter, 20‑degree‑of‑freedom DG‑5F‑S hand aimed at accelerating the commercial integration of dexterous humanoid manipulators.
The Palo Alto, Calif.-based company said it has designed NEO for household use. The humanoid is available through 1X’s early access program for $20,000, which includes priority delivery in 2026. A subscription model will also be available for $499/month.
NEO can do old tasks in unfamiliar places, plus learn new ones
With this update, users can give NEO a simple voice or text prompt, and the robot uses what it’s looking at to generate visualizations of future actions. A built-in inverse dynamics model then translates these into precise movements for NEO to complete the request, 1X explained.
“With the 1X World Model, you can turn any prompt into a fully autonomous robot action — even with tasks and objects NEO’s never seen before,” said Daniel Ho, an AI researcher at 1X. The Robot Report
1X launches world model enabling NEO robot to learn tasks by watching videos
1X said its new world model puts it a step closer to a future where robots can teach themselves to do any task a human can do. | Source: 1X Technologies
1X Technologies AS last week announced its latest 1X World Model. The company said the AI update for NEO enables the humanoid robot to turn any request into an AI capability on demand, using a video model grounded in real-world physics. ## 1X says robot can learn on its own
Where traditional AI models for humanoid robots have relied on data collected by human operators, the 1X World Model enables NEO to collect its own data and autonomously master new capabilities. This opens the door for robots to eventually teach themselves anything, 1X claimed.
Where improvement in AI capabilities for humanoids has been bottlenecked by the speed in which robot data can be collected by human operators, 1X said its World Model doesn’t only self improve from NEO collecting it’s own data. It also benefits from the improvement of video models, given that the world model uses a video model at its core.
Mentioned in this article