Wang Xingxing Says Unitree G1 Has Rolled Off the Line About 10,000–20,000 Units; Physical AI May Break Through in One to Three Years

Report timeline

1. Unitree's Wang Xingxing: The Tipping Point for Physical AI Is Within the Next Year or Two

   Speaking at the 2026 World Physical AI Conference, Wang Xingxing said that GPT-6 can directly control a robot to paint and autonomously wash brushes and change colors with zero training, and that robot planning and navigation problems have been largely solved; the remaining difficulties are the success rate of interaction with real objects and alignment with the physical world. He expects large models to solve assembly problems in one to three years, with voice-commanded room tidying achievable as soon as next year, and announced that the World Physical AI Competition will feature eight tracks and offer prizes of several tens of millions of yuan each year.

   人形大讲堂 · 微信公众号 · Oct 26, 2026

2. Unitree's Wang Xingxing: Physical AI Nearing a Tipping Point, GPT-6 Already Deployed to Unitree Robots With Zero Training

   Speaking at the 2026 World Physical AI Conference, Unitree Technology founder Wang Xingxing said physical AI is showing signs of an imminent breakthrough, likely within the next one to two years, and that GPT-6 has been deployed to Unitree robots with completely zero training, enabling tasks such as drawing on its own, playing board games, and finding kitchen tools.

   智东西 · Oct 26, 2026

3. Unitree Hosts 2026 World Physical AI Conference; Wang Xingxing Says Embodied Intelligence's ChatGPT Moment Could Arrive as Early as 2027

   At the "2026 World Physical AI Conference and Partner Ecosystem Conference" hosted by Unitree Robotics, Wang Xingxing said the ChatGPT moment for embodied intelligence could arrive as early as 2027 or 2028, and proposed a "double 80% standard": completing 80% of tasks in 80% of unfamiliar scenarios using only voice or text instructions.

   差评X.PIN · 微信公众号 · Oct 24, 2026

4. Unitree's Wang Xingxing: Physical AI May Break Through Within One to Three Years; Unitree Developing Self-Evolving Model System

   Unitree Technology held the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference in Hangzhou. Founder Wang Xingxing said in a speech that physical AI may see a major breakthrough within one to three years, and said GPT-6 Astra enabled Unitree robots to complete painting and Go-playing experiments under zero-training conditions. He also announced Unitree's in-development Physical AI Robot Self-Evolution V1 system, which lets AI agents write code autonomously, test and train in simulated environments, and validate in real environments, forming a continuous iteration loop; he hopes Unitree robots can enter ordinary households by the end of this year or next year.

   36Kr Embodied Intelligence · Oct 23, 2026

5. Wang Xingxing says embodied AI commercialization still needs several years; industry may enter a period of adjustment

   In a keynote speech at the 2026 World Physical AI Conference, Unitree's Wang Xingxing said physical AI is showing signs of a breakthrough, but it may still take 1-2 or 2-3 years. The core challenge of embodied AI is aligning AI model inputs and outputs with real robots; large models can plan navigation but have not solved problems after physical contact. Commercial services will be deployed faster than home and industrial applications.

   OFweek Robotics · Oct 23, 2026

6. Unitree's Wang Xingxing Exposes Industry Illusion: No Matter How Powerful a Large Model Is, It's Useless If It Can't Align with a Real Robot

   On October 22, the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference was held in Hangzhou. In his keynote speech, Unitree's Wang Xingxing judged that for embodied intelligence to pass the ChatGPT moment, the core challenge is not whether the model is smart enough, but the alignment between the input and output of AI models and real robots.

   OFweek Robotics · Oct 23, 2026

7. Morning Brief | WeChat responds to not being adapted for iPhone Duo / BYD: L3 expected to land in 2028 / Rokid color-display AI glasses accidentally exposed

   Yang Dongsheng, senior vice president of BYD, said at the World Intelligent Connected Vehicles Conference that L3-level conditional autonomous driving is expected to land in 2028, and higher levels may land another 1–2 years after that. Wang Xingxing, founder of Unitree Robotics, said at the 2026 World Physical AI Conference that physical AI is showing signs of being "broken through," with a breakthrough possibly emerging in the next 1–2 or 2–3 years, and that the core challenge of embodied intelligence is aligning AI model inputs and outputs with real robots.

   爱范儿 · Oct 23, 2026

8. Unitree's Wang Xingxing: The World's Faith in AI Is Still Not Enough, Including My Own

   At the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree Technology CEO Wang Xingxing delivered a roughly 30-minute speech, saying that almost everyone in the world still does not place enough importance on and faith in AI, including himself; physical AI has shown signs of being broken through, likely to happen in the next 1 to 2 years or 2 to 3 years, and the industry is moving toward the ChatGPT moment of embodied intelligence, that is, in 80% of unfamiliar scenarios, robots can complete about 80% of tasks according to voice or text instructions.

   CVer · 微信公众号 · Oct 22, 2026

9. Unitree's Wang Xingxing: Physical AI May See Major Breakthrough in One to Three Years; Unitree Developing Self-Evolving Model

   Speaking at the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree founder Wang Xingxing said that with improvements in general large models, physical AI may see a major breakthrough within one to three years. He said that GPT-6 Astra, under zero-training conditions, enabled Unitree robots to complete painting and Go-playing experiments, but he also noted that large models still have clear shortcomings in perceiving changes in the real environment and correcting operational errors.

   36氪 · 微信公众号 · Oct 22, 2026

10. Unitree's 10th Anniversary: Wang Xingxing Says He Didn't Believe in AI Enough

    Unitree Robotics held the World Physical AI Conference in Hangzhou to celebrate its 10th anniversary. In a 36-minute speech, founder Wang Xingxing mainly urged people to believe in AI and reflected that he himself had not believed enough. He shortened his expected timeline for embodied intelligence's ChatGPT moment from 2 to 3 years, or 5 to 10 years, as of August this year to 1 to 2 years, or 2 to 3 years at the slowest.

    极客公园 · Oct 22, 2026

11. Wang Xingxing: GPT-6 Already Deployed on Unitree Robots with Zero Training; Physical AI May Be Cracked Within 1-2 Years

    At the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree Technology's Wang Xingxing gave a roughly 30-minute speech, saying that GPT-6 has been fully deployed on Unitree robots with zero training, enabling them to autonomously paint, clean paintbrushes, play Go and chess, and find kitchens and tools on their own. He judged that physical AI being cracked is very likely to happen in the next 1-2 or 2-3 years, and that the industry is moving toward an embodied-intelligence ChatGPT moment, in which robots can complete about 80% of tasks in 80% of unfamiliar scenarios based on voice or text instructions.

    机器人前瞻 · 微信公众号 · Oct 22, 2026

12. Wang Xingxing at World Physical AI Conference: Physical AI shows signs of being 'broken through,' core challenge is model-robot alignment

    In his keynote at the 2026 World Physical AI Conference, Unitree founder Wang Xingxing said that with the release of GPT-6, physical AI has shown signs of being broken through, and a breakthrough could come within the next 1 to 2 years or 2 to 3 years. He believes the core challenge for embodied intelligence to pass the ChatGPT moment is aligning AI model inputs and outputs with real robots, and judges that commercial service scenarios will be deployed faster than home and industrial ones. According to the on-site introduction, the G1 robot released in May 2024 has cumulatively produced about 10,000 to 20,000 units.

    未来机器人网 · 微信公众号 · Oct 22, 2026

13. Unitree's Wang Xingxing: Embodied intelligence has passed the GPT moment, biggest challenge is aligning models with real robots

    At the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree founder Wang Xingxing said the GPT moment for embodied intelligence means robots can successfully complete about 80% of main tasks in about 80% of unfamiliar scenarios via voice or text commands. He believes large models already have strong capabilities in robot planning and navigation, and that the biggest challenge in passing the GPT moment remains aligning AI model inputs and outputs with real robots. He expects that, at the fastest, by the end of this year or next year, robots may be able to complete tasks such as tidying a room on command.

    36Kr Embodied Intelligence · Oct 22, 2026

14. Unitree's Wang Xingxing: Physical AI Is Showing Signs of Breaking Through

    At the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree founder Wang Xingxing said that with the release of GPT-6, physical AI is showing signs of breaking through and could achieve a breakthrough in the next 1-2 or 2-3 years. He believes the core challenge for embodied intelligence to pass the ChatGPT moment is aligning AI model inputs and outputs with real robots, and that deployment in commercial services will come faster than in home and industrial settings.

    36Kr Embodied Intelligence · Oct 22, 2026

15. Unitree's Wang Xingxing: G1 robot cumulative production reaches about 10,000–20,000 units

    At the 2026 World Physical AI Conference and Unitree Partner Ecosystem Conference, Unitree founder Wang Xingxing said that since its release in May 2024, the Unitree G1 robot has reached cumulative production of about 10,000–20,000 units.

    36Kr Embodied Intelligence · Oct 22, 2026