
Embodied AI
126 related stories

Can VLA Models Finally Do Away with Post-Training? Self Variable's Open-Source WALL-OSS-0.5 Deploys Directly from Pre-training
Self Variable Robotics has open-sourced WALL-OSS-0.5, an Embodied AI foundation model that achieves strong zero-shot generalization on real robots directly after pre-training, challenging the industry norm that VLA models require post-training for deployment.

The First Wave of Consumer Quadruped Robots Enters Homes: What Are These Users Thinking?
As embodied AI products begin to enter households, early adopters share their experiences with 'Big Head BoBo', revealing that emotional value and continuous OTA updates are key drivers for consumer acceptance.

Behind the 100 Billion RMB Revenue Target for 2030, What is Zhiyuan Really Trying to Achieve?
Zhiyuan Robotics outlines its 'XYZ Curve' strategy, aiming for 100 billion RMB in revenue by 2030. The company is shifting from performance-based demos to practical productivity applications, introducing new models like BFM and GCFM, and pioneering a RaaS (Robotics as a Service) business model akin to AWS.

Just Now: Vbot WeiTaDongLi Completes Nearly 500 Million RMB Financing, Targets Humanoid Robot R&D in Next Phase
Vbot WeiTaDongLi has completed a Pre-A round of financing totaling nearly 500 million RMB, marking the largest single investment in the consumer-grade Embodied AI sector to date. The funds will support mass production and the R&D of next-generation humanoid robots.

Opening Doors, Making Beds, and Nodding: Figure Turns Two Robots into 'Colleagues'
Figure AI demonstrates multi-robot collaboration with two Figure 03 units using the Helix 02 model to perform complex bedroom tasks, highlighting advancements in embodied AI and scalable deployment.

How GENE-26.5 Teaches Robots to Crack Eggs, Slice Tomatoes, and Play Piano: The Secret Isn't Just in the Model
Genesis AI releases GENE-26.5, a full-stack Embodied AI system enabling robots to perform complex dexterous tasks like cooking and piano playing, proving that hardware and control systems are as critical as the model itself.

Aiming for 100,000 Robots in 4 Years, Figure Has Delivered 350 So Far
Figure AI has delivered over 350 Figure 03 robots, boosting production capacity 24-fold to one unit per hour. While trailing competitors like Unitree and Agibot in volume, Figure leverages scale to enhance its Helix 02 model through real-world data, setting the stage for intense industry competition in 2026-2027.

PI Bets Everything on OS: Can One Model Adapt to All Robots?
Physical Intelligence's co-founder Quan Vuong discusses the robot 'ChatGPT moment', data bottlenecks, and cloud-based deployment strategies in a recent YC podcast interview.

Xiaomi Open-Sources Embodied Model Post-Training on Real Robots: Can It Stop 'Slacking Off'?
Xiaomi releases the full workflow for post-training its Xiaomi-Robotics-0 embodied model on real hardware, using only 20 hours of data to master fine-grained tasks. The open-source release addresses critical challenges in action continuity and perception robustness.

Optimus V3 Delayed: What Is Musk Planning?
Elon Musk reveals Optimus V3 is ready for movement but delayed to protect IP. Mass production targets July-August, with Fremont and Texas factories scaling up.

What Happened to the Top 3 Humanoid Robots from Last Year's Marathon? Who Will Take the Crown This Year?
Last year's top three humanoid robots in the half-marathon have diverged into distinct paths: industrial applications, consumer-grade companionship, and bionic emotional support. As the 2026 race approaches with stricter autonomy rules, the competition intensifies.

Zhiyuan Kuantuo Launches All-Terrain Intelligent Mobile and Operation Platform, Targeting 300,000 Units Shipped by 2030
Zhiyuan Kuantuo unveiled its all-terrain intelligent mobile and operation platform strategy, aiming for 300,000 units shipped by 2030. The new quadruped robot series leverages full-stack AGI technology to enhance adaptability and autonomy in extreme environments.