Understand this story
On October 29, 2026, Zhiyuan released its new-generation embodied intelligence foundation model GO-3 Preview, along with the AHA3 hardware platform and the industrial robot G2 Max. According to a report by Shangguan News, GO-3 Preview connects the entire process from training and simulation practice to continuous learning after taking up a post, ranking first in three tests: RoboDojo, RoboTwin, and RoboColiseum, and is planned to be fully open-sourced by the end of 2026; in a real-world convenience store scenario test, the Elf G2 robot's task success rate was 28.6% on the first day and rose to 85.7% after five days. QbitAI reported that Zhiyuan also released the omni-modal interaction large model WITA-OMNI and the embodied native operating system Lingqu OS at the same event, and proposed the unit Worken for measuring a robot's effective intelligent working hours. On October 30, Zhiyuan co-founder Yao Maoqing said in an interview with Securities Times that general large models have obvious shortcomings in physical robot operation, and that future systems will move toward a layered architecture; he mentioned that GO-3 Preview adopts the WRAM native architecture, and its zero-shot average success rate on 100 tasks across 20 skill categories increased from 13.4% to 56.4%.
AI-generated from public reports
Developments
2 confirmed developments
- Source:English text is being prepared.
- Source:Shangguan News
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Oct 30
The Beijing News
After 20,000 units rolled off the line, AgiBot proposes Worken and releases GO-3 Preview, aiming to achieve a robot-to-engineer ratio of 20:1
After its 20,000th robot rolled off the production line, at the 2026 AgiBot Innovation Day, AgiBot proposed Worken, a unit for measuring a robot's effective intelligent working hours, and released the heavy-duty embodied robot Genie G2 Max, the embodied foundation model GO-3 Preview, and the omni-modal interaction large model WITA-OMNI. AgiBot disclosed that in the first half of this year it shipped 9,700 robots and generated revenue of nearly 2 billion yuan, and it plans to raise the robot-to-engineer ratio from 1:1 to 10:1 by the end of this year and 20:1 next year. Peng Zhihui said that only when customers shift from asking "can it be built" to "can it be afforded to maintain" will the industry truly confront the full lifecycle of robots.
The Beijing NewsRelated discussion
The Beijing News joined this discussion
This post contributes to heat; it is not a formal report or confirmed fact.
QbitAI
Zhiyuan releases GO-3 Preview and WITA-OMNI dual models, and proposes the Worken work-hour unit
At its Innovation Day, Zhiyuan released the embodied foundation model GO-3 Preview and the omni-modal interaction large model WITA-OMNI; the former topped three benchmarks—RoboDojo, RoboTwin, and RoboColiseum—while the latter ranked first on four benchmarks: Daily-Omni, TimeLens-Bench, EmbSpatialBench, and MMAU. It also launched the third-generation hardware architecture AHA3, the heavy-duty robot Jingling G2 Max, and the embodied native operating system Lingqu OS, and plans to fully open-source Lingqu OS 1.0 by the end of the year. Zhiyuan also proposed Worken, a unit for measuring a robot's effective intelligent work hours, with the formula being the number of active robots multiplied by autonomous working duration multiplied by the output efficiency ratio.
Yicai
Zhiyuan Proposes Worken Effective Intelligent Work Hours, Robot Billing Model Still Awaits Validation
At its Innovation Day on October 29, Zhiyuan proposed the concept of Worken (effective intelligent work hours), using "number of active robots × autonomous working hours × output efficiency ratio" to measure the actual output of autonomous robot operations, and has already calculated Worken for scenarios including the Nanchang Longcheer production line, Guangzhou Metro, and Zhuhai Chimelong, among which the 300 robots at Chimelong reached 33,480 Worken. This billing method is currently conducted only internally at Zhiyuan and has not yet formed a business model; the company is applying for the Worken trademark. Peng Zhihui also released new products including GO-3 Preview and WITA Omni1.0, and stated that in the first half of this year Zhiyuan ranked first globally with 9,700 units shipped and a share of over 43%.
Sina Tech
Zhiyuan Unveils GO-3 Preview Foundation Model, Genie Master Framework, and Genie G2 Max Robot
At the Embodied Work Productivity sub-forum of the 2026 Zhiyuan Innovation Day, Zhiyuan unveiled the embodied foundation model GO-3 Preview, the embodied Agent framework Genie Master, and the heavy-load embodied industrial robot Genie G2 Max. GO-3 Preview adopts the unified WRAM architecture for understanding, prediction, and action generation; on 100 tasks across 20 skill categories, its Zero-Shot average success rate rose from 13.4% to 56.4%, and in model-scale expansion experiments it rose from 56.2% to 73.2%; it has also topped the three major evaluation leaderboards RoboDojo, RoboTwin, and RoboColiseum. Co-President Yao Maoqing proposed three capability standards—measurable, replicable, and sustainable—and the event also presented data including a 98.3% success rate on a phone packaging task and the Genie G2's convenience-store task success rate rising from 28.6% on the first day to 85.7% by the fifth day.
Securities Times
Zhiyuan's Yao Maoqing Says General Large Models Struggle with Physical Operations; Embodied Foundation Models Will Move Toward Layered Architecture
Yao Maoqing, co-founder of Zhiyuan, said in an interview with Securities Times that general large models have obvious shortcomings in physical robot operations, and future systems will move toward a layered architecture: general large models handle high-level long-horizon planning, world models and VLA output mid-level skills, and the low level is handled by a 100–1000Hz high-frequency motor real-time control layer. He pointed out that a general large model may need 1–2 hours to complete a physical operation and consume a large number of tokens, and that in real-world tests its success rate on tasks involving highly dynamic motion and force/tactile feedback is almost zero. At the 2026 Zhiyuan Innovation Day, Zhiyuan released its new-generation embodied foundation model GO-3 Preview, which adopts the WRAM native architecture and is planned for full open source by the end of 2026; on 100 tasks across 20 skill categories, the model's average zero-shot success rate rose from 13.4% to 56.4%.
Securities Times
Zhiyuan's Peng Zhihui: 10,000 Humanoid Robots Rolling Off the Line Is Only a Manufacturing Victory; 10,000 Robots Truly "On the Job" Is the Victory of Physical AI
Peng Zhihui, co-founder, president and CTO of Zhiyuan, said at the 2026 Zhiyuan Innovation Day that 10,000 humanoid robots rolling off the production line is only a victory for manufacturing, and that 10,000 robots truly "on the job" is the victory of physical AI. Zhiyuan also released a measurement unit, Worken (effective intelligent work hours), to measure the actual output of robots autonomously and continuously completing valuable work, and hopes to bill by Worken work hours in the future. Peng Zhihui divided the development of physical AI into five stages, believing that Zhiyuan has passed the first stage and is moving toward deployment in vertical-domain scenarios, while the home is the endpoint rather than the starting point.
Shanghai Observer
Zhiyuan Releases Embodied Model GO-3 Preview, Plans Full Open Source by End of 2026
Zhiyuan has released its new-generation robot embodied model GO-3 Preview, which connects the full pipeline from training and simulation practice to continuous learning after deployment, ranking first in three tests: RoboDojo, RoboTwin, and RoboColiseum, with plans for full open source by the end of 2026. In real convenience store scenario tests, the Elf G2 robot's task success rate was 28.6% on the first day and rose to 85.7% after five days; for the small hook hanging task, the success rate was 17.7 percentage points higher than with traditional training methods.
English text is being prepared.
AgiBot Releases GO-3 Preview Embodied Foundation Model, First to Propose the Worken Workload Metric
AgiBot has released its new-generation embodied foundation model GO-3 Preview, which adopts its self-developed WRAM architecture to integrate task understanding, future state prediction, and action generation into a unified model, and is the first to propose the Worken evaluation framework for measuring a robot's actual work output. AgiBot says the model ranks first overall in three embodied intelligence benchmarks: RoboDojo, RoboTwin, and RoboColiseum. Worken equals the product of the number of active robots, autonomous working hours, and the output efficiency ratio; AgiBot also released the AHA3 hardware platform and Lingqu OS 1.0 at the same time.
Wallstreetcn
Zhiyuan's Peng Zhihui responds to GPT-6 Astra: the stronger the large model, the more important the robot body
At Zhiyuan's fourth Innovation Day, co-founder Peng Zhihui (Zhihui Jun) responded to the question raised by GPT-6 Astra's ability to directly control robots, arguing that regardless of whether foundation models can handle embodied capabilities across the board, body manufacturers that can scale, keep costs low, and achieve high consistency will be more advantaged. He cited data showing Zhiyuan shipped more than 5,000 units in 2025 and 9,700 units in the first half of 2026, and said that networked post-training with 100 real machines can raise task success rates from about 50% to 98%. Zhiyuan also proposed using Worken to measure a robot's effective intelligent work hours, and pointed out that the human-to-machine ratio in operations and maintenance is a key bottleneck for scaling.
The Paper
Zhiyuan's Peng Zhihui: The full-stack route is a mandatory choice for humanoid robot companies
Peng Zhihui, co-founder, president and CTO of Zhiyuan, said at the 2026 Zhiyuan Innovation Day that the full-stack route is a mandatory choice rather than an option for humanoid robot companies. He believes that being only a hardware body supplier would leave a company with weak competitiveness; it must be scenario-oriented and combine sufficiently generalized model capabilities, a reliable body, and software-hardware system capabilities to meet deployment-state requirements.
The Paper
Zhiyuan's Peng Zhihui: Physical AI Still in Early Stage, Proposes Worken Effective Intelligent Work-Hour Framework
At the 2026 Zhiyuan Innovation Day, Peng Zhihui, co-founder, president and CTO of Zhiyuan, said that physical AI is still in an early stage; Zhiyuan has passed the first stage of possessing a stable, low-cost robot body and is advancing the second stage of deployment in vertical segmented scenarios. He proposed the Worken (effective intelligent work-hour) measurement framework, with the formula: number of active robots × autonomous working hours × output efficiency ratio, hoping to bill by work-hours in the future. In the first half of 2026, Zhiyuan's revenue was close to 2 billion yuan and it delivered 9,700 robots; in September, its cumulative output of general-purpose AI robots reached 20,000 units. On the same day, it released the embodied intelligence foundation model GO-3 Preview, the AHA3 hardware platform and the industrial robot G2 Max.
Oct 29
The Paper
Zhiyuan's revenue in the first half of 2026 was nearly 2 billion yuan; Peng Zhihui proposes five leaps of physical AI
Zhiyuan co-founder, president and CTO Peng Zhihui revealed at the 2026 Zhiyuan Innovation Day launch event that in the first half of this year, cumulative off-line output of Zhiyuan general-purpose AI robots reached 20,000 units, revenue approached 2 billion yuan, and the company captured a 43% share of the global humanoid robot market. Peng Zhihui also released the physical AI value metric Worken (effective intelligent work hours), with a formula of the number of active robots × autonomous working hours × output efficiency ratio, which has already been deployed in scenarios such as visitor services in Chengdu's Kuanzhai Alley, loading and unloading on Longqi production lines, and subway security-check guidance. He also proposed that physical AI will undergo five leaps from usable to general-purpose, from a usable body, deployment in segmented scenarios, a generalized embodied foundation model, and swarm intelligence, ultimately moving toward intelligence self-bootstrapping.
Caixin
Zhiyuan's First-Half Revenue Nears 2 Billion Yuan, with 9,700 Robots Shipped
Peng Zhihui, co-founder, president and CTO of Zhiyuan, said at the 2026 Zhiyuan Innovation Day that in the first half of 2026, Zhiyuan's robot shipments reached 9,700 units and revenue approached 2 billion yuan; as of September, a cumulative 20,000 robots had rolled off the production line. Zhiyuan also stated that the penetration of general foundation models into the embodied field is a major trend, embodied models will become increasingly large, and this year's training cluster will approach a scale of 10,000 GPUs. A day earlier, Unitree Technology released its first financial report since going public, with cumulative revenue of 1.817 billion yuan in the first three quarters.
Heat trend for this story
- Current
- 5
- Period peak
- 73 Oct 31 02:00
- 24-hour change
- -87%
Move the pointer or select the chart to inspect hourly heat. Use the left and right arrow keys with a keyboard.