Unitree G1 Timeline

Unitree G1 units were deployed at Tokyo Haneda Airport in a partnership with Japan Airlines, tasked with baggage-handling operations. The deployment marked one of the first commercial airport applications for an affordable consumer-tier humanoid robot, and was cited by Unitree as evidence of the G1's readiness for real-world logistics work.

Unitree Robotics filed for a domestic A-share IPO on a Chinese stock exchange, targeting a valuation of approximately $580 million. The company reported 335% revenue growth year-over-year in 2025, driven largely by G1 unit shipments. The IPO filing signals Unitree's intent to scale manufacturing capacity to between 10,000 and 20,000 G1 units in 2026.

A Unitree G1 completed a 130,000-step outdoor walk in Altay, Xinjiang, at temperatures as low as −47.4 °C, tracing an Olympic-ring emblem across a 186 m × 100 m area using Beidou satellite positioning. The demonstration was staged ahead of the 2026 Winter Olympics and highlighted the G1's cold-weather operational resilience.

Multiple Unitree G1 units performed parkour sequences and kung fu routines at the 2026 Chinese New Year Spring Festival Gala, broadcast nationally to hundreds of millions of viewers. The performance built on the G1's appearance at the previous year's gala and reinforced Unitree's position as China's most commercially successful humanoid robot manufacturer.

Unitree confirmed that its 2025 factory output exceeded 6,500 total units across all robot lines, with more than 5,500 of those being G1 humanoids. The figures significantly outpaced US-based humanoid manufacturers in unit volume and established the G1 as one of the best-selling humanoid research platforms of 2025 by shipments.

Unitree introduced the G1-D, a wheeled-base variant of the G1 that mounts the humanoid upper body on a wheeled mobile chassis rather than bipedal legs. The G1-D carries 17–19 upper-body DoF, a top speed of 1.5 m/s, and an 80 kg payload capacity on the chassis. It was positioned for industrial and logistics applications where wheeled mobility offers practical advantages over bipedal walking.

G1 EDU platforms were adopted by research groups at MIT, Stanford, Carnegie Mellon University, ETH Zurich, and Tsinghua University, among others, making the G1 one of the most widely used humanoid research platforms in academic settings. Tsinghua University demonstrated G1 playing table tennis using the LATENT reinforcement learning framework. Unitree also released UnifoLM-VLA-0, an open-source vision-language-action model providing manipulation baselines across 12 task categories.

G1 units performed traditional yangge dance at the 2025 Chinese New Year Spring Festival Gala, gaining widespread public attention in China. Following the broadcast, Unitree made the G1 available for purchase on the JD.com e-commerce platform on 12 February 2025, broadening its domestic retail availability beyond the Unitree store.

On 16 January 2025, Unitree announced the latest upgrade plans for the G1, focused primarily on the robot's walking and running gaits. The update addressed stability at higher speeds and improved energy efficiency across the locomotion cycle, reflecting ongoing software development alongside the hardware production ramp.

Unitree began mass production of the G1 on 20 August 2024, approximately three months after the ICRA launch. Production ramped through the second half of the year, with deliveries to early buyers and research institutions beginning in Q3 2024. The quick transition from announcement to mass production was notable compared to other humanoid robot manufacturers.

Unitree unveiled the G1 at the 2024 IEEE International Conference on Robotics and Automation (ICRA) in Yokohama, Japan, on 13 May 2024. The robot was introduced as a "Humanoid Agent AI Avatar" and priced from US$16,000 for the base configuration, substantially below any comparable bipedal humanoid then available. Unitree CEO Wang Xingxing presented the platform at ICRA seminars on 17 May 2024. The G1 was announced alongside the G1 EDU tier, aimed at developers requiring full SDK access and expanded hardware options.