As another productive AI breakthrough for the implementation of the “General-Purpose Warehouse Robot” strategy, Gino 1 distinguishes itself from other humanoid robots. It is purpose-built for warehousing, designed and trained exclusively for general multi-task capabilities in warehouse scenarios.
Its embodied brain, Geek+ Brain, deeply integrates massive warehousing data accumulated by Geek+ over years, combined with large-scale simulation and reinforcement learning. This empowers Gino 1 with human-like general manipulation abilities, enabling it to perform warehousing tasks such as picking, carton handling, packing, and inspection. It truly realizes “one robot covering mainstream manual operation scenarios in the warehouse”, leading a critical leap in the warehousing industry from mobility intelligence to operational intelligence.
Beyond technological breakthroughs, Geek+ also focuses on the commercial essence of general-purpose robots. The company is committed to enabling general-purpose robots to create commercial value in real-world scenarios. Through scenario-oriented design and targeted data training, Gino 1 not only meets customer requirements in reliability and efficiency but also delivers outstanding cost-effectiveness, making commercial deployment more attractive to customers and accelerating real-world application.At present, the product has mature mass-production capacity and can be quickly deployed on a large scale, laying a foundation for the widespread implementation of full-process unmanned warehouses.
The launch of Gino 1 marks a key milestone in Geek+’s innovation journey: from dedicated warehouse robots (AMR) in the “goods-to-person” model, to dedicated robotic arm unmanned picking stations for high-efficiency selection of massive SKUs, to today’s general-purpose warehouse robot Gino 1. Geek+ has successfully built an intelligent warehousing operation system featuring multi-agent full-process collaboration, fully leading a new era of unmanned warehouses.
Scene-Native
Productive AI with Built-In Know-How
Unlike the industry approach of “developing technology first and then finding scenarios”, Gino 1 is a “warehouse-native” embodied intelligence solution created by Geek+, backed by the world’s No.1 warehouse fulfillment market share and a vast customer and channel network.It directly inherits Geek+’s decade of scenario know-how and industry practice, forming unique technological differentiation and greatly accelerating commercialization.
Full-Stack Self-Developed Hardware
Enabling the robot to move smoothly, grasp stably, and operate reliably
Gino 1’s hardware system is fully optimized for warehouse pain points and deeply aligned with the entire warehousing operation workflow:
Head Multi-Eye Vision
The robot’s head is equipped with triple main vision and front/rear fisheye cameras, supporting high-precision close-range recognition and 360° environmental perception and semantic understanding, ensuring accuracy and safety for multi-task complex operations.
Three-Finger Dexterous Hand
The robot arm features a 7-active-degree-of-freedom three-finger dexterous hand, balancing grasping flexibility and reliability. It is also equipped with a palm binocular camera and tactile sensors, enabling multi-dimensional operational perception.
Humanoid Dual Arms with Full-Joint Force Control
The dual arms have 14 degrees of freedom in a humanoid structure for enhanced flexibility. Full-joint force control ensures safe interaction with humans and the environment. The maximum payload of 20 kg per arm covers general warehousing load requirements.
Ultra-Long Operation Endurance
In addition to a 50Ah high-capacity battery, the robot features an industry-first wireless super-fast charging function with 3KW charging power, greatly improving charging efficiency and extending working hours.
3D Environmental Perception
Based on Geek+’s decade of AMR R&D in intelligent environmental perception, positioning, and navigation, the chassis is equipped with front and rear hemispherical 3D LiDARs, enabling 3D scene reconstruction, dynamic trajectory planning, and flexible omnidirectional movement.
Super Computing Power
The main controller uses the NVIDIA Jetson Thor robotic computer based on NVIDIA Blackwell, delivering up to 2,070 FP4 TFLOPS of computing performance. It efficiently handles parallel computing of multi-modal AI models, unlocking complex scenarios such as physical AI, high-speed sensor data processing, and general robot tasks, providing strong computing support for the robot brain.
Embodied Intelligence VLA Fast-Slow Collaborative System
Enabling the robot to truly understand, think clearly, and act accurately
Gino 1’s software system deeply integrates warehousing scenario know-how with AI technology to create a new-generation embodied intelligence VLA fast-slow collaborative system.It allows the robot to imitate human cognitive and execution capabilities, perform fully autonomous operations throughout the warehouse, and significantly improve the stability and success rate of complex tasks.
Core Technical Capabilities of Geek+ Embodied Intelligence VLA System:
- Multi-Modal Fusion Perception: Integrates vision, language, and proprioception to accurately identify goods/spatial relationships, parse task instructions, and capture self-posture, adapting to complex and dynamic warehouse conditions.
- Fast-Slow Collaborative Intelligence Architecture: The slow system handles high-level task understanding and planning (instruction decomposition, global strategy generation); the fast system performs real-time visual perception and feedback control (target tracking, rapid obstacle avoidance), balancing rational decision-making and efficient execution.
- High-Frequency Closed-Loop Action Decision Center: Unified action generation model with high-frequency command output, realizing visual-force closed-loop control to ensure precise operation and stable movement for reliable performance in complex scenarios.
- Data-Driven Continuous Evolution: Based on massive warehousing operation data and efficient data closed-loop capabilities, the algorithm model is continuously optimized to improve scenario adaptability and task execution for diverse warehousing needs.
With these software capabilities, Gino 1 brings four core values to warehouse automation:
Efficient Execution of Complex TasksStably supports multi-step, multi-target continuous workflows to meet diverse and flexible operational demands in warehouses.
Interpretable & Manageable Intelligent BehaviorTransparent and controllable task planning and execution logic for easy on-site operation, system tuning, and large-scale management.
Industrial-Grade Operational StabilityHigh-frequency closed-loop control and real-time feedback ensure precise operation and reliable movement, meeting long-term industrial operation requirements.
Scenario Continuous Evolution CapabilityData-driven algorithm iteration makes the robot more stable and adaptable over time, continuously improving generalization and operational performance.
According to Interact Analysis, over 70% of warehouses worldwide still heavily rely on manual labor. Even in modern warehouses equipped with AGV/AMR, operational tasks such as picking and sorting — which account for more than 50% of operating costs — have long been an automation “blind spot” due to complex movements and high flexibility requirements.This massive market waiting to shift from “human hands” to “robotic hands” contains hundreds of billions of dollars in commercial value.The launch of Gino 1 precisely fills this gap, strongly opening up huge growth space for “operational intelligence”.
Full-Domain Collaboration
Implementable Full-Process Unmanned Warehouse Solution
Based on in-depth scenario understanding, Geek+ has built a clear and implementable full-process unmanned warehouse solution:Relying on the Geek+ Brain embodied intelligence foundation model, the AMR cluster completes mobile tasks such as transportation and storage, while Gino 1 fully takes over flexible and complex manual operation scenarios. For high-efficiency picking of massive SKUs, picking efficiency can be further enhanced by unmanned picking stations.
This solution achieves full intelligent coverage of core warehouse operations, greatly easing labor pressure for customers, reducing long-term costs, and achieving a leap in efficiency.
Geek+ has completed the first phase “single-point breakthrough” of its “General-Purpose Warehouse Robot” strategy:the unmanned picking station dedicated to picking scenarios launched in late October 2025 has recently been accepted by international giants, completing technical and commercial validation.The release of Gino 1 marks the delivery of the second phase strategy “scenario generalization”, realizing the evolution from dedicated to general-purpose robots.
Going forward, Geek+ will accelerate the large-scale implementation of unmanned warehouse solutions through its global network, and promote dual evolution of technology and commerce via the closed-loop iteration of “scene – data – technology”.Meanwhile, the company will prepare to launch the third strategic phase of ecological co-construction, joining hands with global physical AI and embodied intelligence enterprises to jointly drive warehousing operations into a new truly unmanned era of multi-agent full-domain collaboration.
