Target Audience and Intended Roles
LimX Oli is an embodied AI research humanoid developed by LimX Dynamics for university laboratories, robotics developers, and commercial system integrators. Standing 165 cm tall, the machine provides a life-size bipedal embodiment for researchers who study whole-body balance, legged locomotion algorithms, and manipulation tasks. LimX Dynamics engineered the chassis to let teams evaluate reinforcement learning models on physical hardware without building custom mechanical structures from scratch.
The machine does not serve heavy industrial manufacturing or automated warehouse picking. A 3 kg payload per arm limits the robot to handling lightweight tools, electronic components, or educational props. Facilities requiring heavy crate transport or continuous high-load pallet manipulation should look beyond research bipeds, and prospective buyers evaluating commercial mobility can consult our Chinese humanoid robots buying guide.
A development pipeline focusing on autonomous domestic chores also falls outside current operating capabilities. The onboard battery provides approximately 1.5 hours of operating runtime, requiring regular manual intervention for battery recharging. Operational teams must deploy LimX Oli in controlled testing spaces where engineers can maintain supervised power management and software safety stops.
Published Physical and Mechanical Specifications
LimX Dynamics' spec sheet states that LimX Oli measures 165 cm (5 feet 5 inches) in height, 55 cm in shoulder width, 70 cm in arm length, and weighs 55 kg or less. The kinematic configuration displays a minor discrepancy across official documentation. The primary LimX Dynamics product page cites 31 degrees of freedom (DoF), allocating 2 to the neck, 7 to each arm, 3 to the waist, and 6 to each leg. The spec sheet lists 33 active DoF; the product page's 31 excludes end effectors.
Peak joint torque is 150 newton-metres (N·m). The bipedal chassis achieves a maximum travel speed of 5 km/h over level ground, which permits walking tests at standard human walking paces. The arms support a maximum single-arm payload of 3 kg (6.6 lb), suitable for sensor placement, small item retrieval, and test manipulation.
Perception hardware combines embedded depth cameras on the head and chest with open physical expansion ports for auxiliary sensors. Engineers can mount external light detection and ranging (LiDAR) modules or wrist-mounted visual sensors via onboard universal serial bus (USB) and Ethernet ports. For end-effector hardware, LimX Dynamics offers an optional 1-DoF two-finger gripper or a 6-DoF five-finger dexterous hand to match specific manipulation projects.
| Spec | Oli (spec sheet) |
|---|---|
| Height | 165 cm (5 feet 5 inches) |
| Weight | ≤55 kg (121.3 lb) |
| Shoulder width | 55 cm |
| Arm length | 70 cm |
| Active degrees of freedom (DoF) | 33 active DoF (31 DoF excluding end effectors on product page) |
| Peak joint torque | 150 N·m |
| Max single-arm load | 3 kg (6.6 lb) |
| Max travel speed | 5 km/h |
| Battery capacity and runtime | 9,500 mAh, about 1.5 hours |
| Motion control compute | Rockchip RK3588 (8 GB RAM, 64 GB storage) |
| Perception compute | NVIDIA Jetson Orin NX (157 TOPS, 16 GB RAM, 1 TB storage) |
| Integrated sensors | Self-developed 6-axis IMU, head and chest depth cameras |
| Optional end effectors | 1-DoF 2-finger gripper or 6-DoF 5-finger dexterous hand |
LimX Oli Pricing and Model Variations
LimX Dynamics introduced LimX Oli on 31 July 2025 with an announced baseline starting price of RMB 158,000, which converted to approximately $21,800 at launch. Launch reporting identified three separate configuration tiers designated as Lite, EDU, and Super. At the time of writing, the official LimX Dynamics technical specification page groups the lineup into two variant names: Oli and Oli EDU. Both variants carry identical base physical dimensions, motor torque ratings, and compute hardware on the published specification table.
Individual tier prices for separate configurations remain unlisted on public portals. LimX's site has no cart. Instead, procurement links direct corporate and university buyers through an enterprise order consultation request to determine final hardware and delivery expenses. Buyers tracking procurement budgets across alternative domestic models can compare figures in our analysis of Unitree G1 price structures.
LimX Dynamics secured $200 million in Series B financing in February 2026, followed by a $200 million pre-IPO round announced on its site in July 2026. Prospective buyers must confirm export handling, warranty coverage and compute tier through LimX Dynamics' order consultation; its published figures say nothing about overseas shipping duties.
Developer Software and Computing Architecture
Oli runs two computers. Perception and locomotion control each get their own board. Real-time balance and limb coordination run on a dedicated Rockchip RK3588 processor equipped with 8 GB of random-access memory (RAM) and 64 GB of onboard storage. Spatial perception and computer vision models execute on an NVIDIA Jetson Orin NX system-on-module providing up to 157 tera-operations per second (TOPS) of compute, 16 GB of memory, and 1 TB of storage.
Software developers interact with the machine using an open Python software development kit (SDK) supporting both high-level behavioral commands and low-level motor joint control. The software framework provides open application programming interfaces (APIs) for controlling limb trajectories, reading the self-developed 6-axis inertial measurement unit (IMU), and capturing video streams from head- and chest-mounted depth sensors.
Simulation pipelines benefit from a complete Unified Robot Description Format (URDF) file published by LimX Dynamics. Roboticists can import the kinematic definition directly into industry-standard physics simulators, including NVIDIA Isaac Sim, MuJoCo, and Gazebo. This simulation parity lets engineering teams train locomotion policies in virtual environments before validating control scripts across the physical hardware platform.
Payload Limits and Battery Endurance Realities
Two numbers set Oli's limits: runtime and arm load. The 9,500 milliampere-hour (mAh) internal battery yields an estimated runtime of about 1.5 hours of continuous activity. For university labs and industrial research teams running full-day experimental protocols, this window requires structured test schedules interspersed with routine recharging cycles.
Arm payload limits are similarly bounded by the 3 kg single-arm maximum capacity. While sufficient for manipulating light objects, handling measurement instruments, or picking lightweight lab samples, this rating rules out real-world freight loading, automotive component assembly, or moving loaded storage totes. Teams requiring heavy dual-arm manipulation need higher payload thresholds than standard research biped humanoids provide.
Mechanical joint limits define the physical operating envelope during dynamic maneuvers. Peak joint torque is 150 N·m and top speed is 5 km/h. LimX publishes no ingress rating or outdoor operating spec, so treat Oli as an indoor lab machine until it does.
Comparing Oli Against TRON 2 and Competing Humanoids
Within the internal LimX Dynamics product catalog, LimX Oli serves a different research function than modular mobile manipulators. The LimX TRON 2 features a modular tri-form architecture that switches between dual-arm, bipedal (sole) and wheeled-leg configurations. TRON 2 provides a dual-arm payload of 10 kg (5 kg per arm) and up to 30 kg in wheeled-leg mode, outperforming Oli on pure payload capacity while sacrificing the anthropomorphic full-body geometry of a 165 cm biped.
Researchers seeking bipedal gait evaluation alongside modular mobility can examine our detailed breakdown in the LimX TRON 1 vs TRON 2 comparison. TRON 2 is the reconfigurable manipulator that climbs 20 cm steps. Oli keeps a human silhouette with a 3-DoF waist, which suits studies of human-like motion planning.
In the broader commercial market, researchers often cross-shop LimX Oli against the Unitree G1 bipedal platform. Both are sold to research labs with open development interfaces. The choice rests on payload, compute integration and the support terms each vendor offers in your territory.
Procurement Evaluation and Hardware Verdict
LimX Oli represents a capable physical hardware platform for academic departments and development teams focused on embodied AI locomotion and vision-based manipulation. Its open Python SDK, dual-compute architecture featuring NVIDIA Jetson Orin NX, and published URDF file remove common development hurdles in physical robotics simulation. It launched at RMB 158,000.
Organizations seeking turnkey industrial utility, automated battery swapping docks, or warehouse logistics automation should not purchase LimX Oli. The 1.5-hour runtime and 3 kg single-arm payload envelope require steady engineer supervision and prevent autonomous round-the-clock physical labor. The verdict would change if LimX released higher-payload arm actuators, a larger battery option or certified self-docking.
Engineering teams with existing budgets for custom battery charging workflows and software development will find the mechanical specifications well-matched to research objectives. Prospective enterprise buyers should use LimX Dynamics' order consultation on limxdynamics.com to confirm current delivery schedules, regional warranty provisions, and accessory pricing.
Bottom Line
The LimX Oli is an embodied AI research humanoid offering an accessible entry point for laboratories exploring bipedal locomotion and lightweight robotic manipulation. It is not suitable for heavy warehouse logistics, factory material handling, or unsupervised domestic duties due to its 3 kg arm payload ceiling and 1.5-hour operating runtime. A hardware revision introducing high-capacity battery packs, automated charging stations, or higher-torque arm actuators would widen its utility beyond supervised academic research. Teams needing an open-architecture research biped should request formal hardware pricing from LimX Dynamics sales representatives.
Explore our comprehensive Chinese humanoid robots buying guide to assess how LimX Oli compares against alternative research and commercial bipedal platforms.
FAQs
How much does the LimX Oli cost?
LimX Oli launched with a starting price of RMB 158,000, which was approximately $21,800 at launch. LimX Dynamics does not publish individual variant checkouts on its website at the time of writing. Commercial and institutional buyers must submit an order consultation on LimX Dynamics' site to obtain current configuration pricing.
How tall is LimX Oli?
LimX Oli stands 165 cm tall, which equals 5 feet 5 inches. The robot measures 55 cm across the shoulders, features an arm length of 70 cm, and has a total operating weight of 55 kg or less according to official technical specifications.
What is the difference between Oli and Oli EDU?
At launch, reports referenced Lite, EDU, and Super tiers, but the current official spec sheet lists Oli and Oli EDU side-by-side with identical physical dimensions, compute components, and joint torque ratings. Specific differences in preloaded software licenses or educational toolkits must be confirmed directly through LimX Dynamics.
Does LimX Oli have hands?
Hands are optional on LimX Oli: the spec sheet lists a 1-DoF two-finger gripper or a 6-DoF five-finger dexterous hand. The official specification sheet details an optional 1-DoF two-finger gripper and an optional 6-DoF five-finger dexterous hand. Buyers configure these end effectors depending on whether their research requires simple clamping or complex multi-finger manipulation.
How long does the Oli battery last?
The LimX Oli battery provides an operating runtime of about 1.5 hours. The system is equipped with a 9,500 mAh battery pack. Because testing sessions require battery replenishment after 90 minutes of active operation, research teams must plan experiment schedules around periodic recharging intervals.
Is LimX Oli open source?
LimX Oli is not fully open-source hardware, but it provides open developer software interfaces. LimX Dynamics supplies an open Python SDK, open high-level and low-level APIs, and a complete URDF file for simulation across NVIDIA Isaac Sim, MuJoCo, and Gazebo environments.