Astribot is a humanoid robot brand developed by Stardust Intelligence (Chinese: 星尘智能), a robotics startup headquartered in Shenzhen, China. Founded in late 2022 by Lai Jie, a former lead at Tencent's Robotics X laboratory, the company gained widespread attention in April 2024 when a demonstration video of its Astribot S1 robot went viral, showcasing remarkable speed and dexterity in everyday manipulation tasks such as pouring drinks, flipping food in a frying pan, and writing calligraphy. The S1 officially debuted at the 2024 World Robot Conference in Beijing on August 21, 2024, and entered limited commercial availability in late 2025.
Astribot's approach centers on a proprietary software-hardware integrated architecture called "Design for AI" (DFAI), combined with imitation learning techniques that allow the robot to acquire new skills by observing human demonstrations. The company positions its products at the intersection of household assistance, research, and commercial service, aiming to bring general-purpose robot assistants to billions of people.
Stardust Intelligence was established at the end of 2022 in Shenzhen, Guangdong Province, China. The company's core founding team consists of six members, all formerly from Tencent's Robotics X laboratory. The founder and CEO, Lai Jie, brings over ten years of experience in robot research and development. He was the first employee (Employee No. 1) at Tencent's Robotics Lab, where he led the development of Tencent's wheel-legged robot called Ollie, among other projects. The Ollie robot attracted attention in the research community for its ability to perform acrobatic backflips using a tail-like mechanism, and the work produced academic publications in journals such as Frontiers in Neurorobotics. Before joining Tencent, Lai Jie served as team leader for Baidu's "Xiaodu Robot" project, an early Chinese conversational robot initiative. He also conducted research at Hong Kong Polytechnic University and holds a master's degree from Wuyi University.
Beyond the founding team from Tencent Robotics X, Stardust Intelligence has recruited engineers with backgrounds at Baidu, Huawei, Google, and UBTECH Robotics, another Shenzhen-based humanoid robot maker. This mix of talent from both major Chinese technology companies and international firms reflects the competitive recruitment environment in China's robotics sector.
The name "Astribot" derives from the Latin proverb "Ad astra per aspera," which translates roughly to "through hardship to the stars." The company interprets this as "a journey through hardship to reach stardust," reflecting its long-term commitment to developing and popularizing AI robot technology. The parent company name, Stardust Intelligence (星尘智能), carries the same thematic reference in Chinese.
Astribot has raised capital through multiple funding rounds, though specific dollar amounts have not been fully disclosed in most cases:
| Round | Date | Key Investors | Notes |
|---|---|---|---|
| Pre-A | Mid-2024 | Matrix Partners, Yunqi Capital, Daotong Investment, Qinghui Investment | Described as "tens of millions of dollars"; Huaxing Capital served as financial advisor |
| Series A | April 10, 2025 | Ant Group, Jinqiu Capital, Matrix Partners | Undisclosed amount; returning investors included Matrix Partners |
As of early 2026, Astribot has six known institutional investors: Matrix Partners, Ant Group, Yunqi Partners, Dalton Venture, Jinqiu Capital, and MPC. The involvement of Ant Group, the fintech affiliate of Alibaba Group, is notable as it signals interest from one of China's largest technology conglomerates in the humanoid robotics space. The company has stated that funding is directed toward top talent recruitment, research and development investment, and commercial deployment of its robots.
The Astribot S1 is the company's flagship product: a mobile dual-arm humanoid robot designed to perform household chores, research tasks, and commercial service operations. Stardust Intelligence reportedly developed the S1 in approximately one year after the company's founding.
On April 29, 2024, Astribot released a demonstration video that rapidly circulated across social media and technology news outlets. The video showed the S1 performing a series of tasks with speed and precision that surprised observers in the robotics community. Publications including New Atlas, Interesting Engineering, Fox News, and Tech Times covered the video extensively. Demonstrated capabilities included:
Notably, when the S1 was first shown in April 2024, it was presented only from the waist up, without revealing its lower body or locomotion system. The video drew comparisons to demonstrations by Boston Dynamics, Figure AI, and other leading humanoid robot developers, though some commentators noted that the lack of a visible lower half raised questions about the platform's completeness at that stage. The publication Maginative ran the headline "Stardust Intelligence's Impressive S1 Robot Demo Raises Eyebrows," noting both the remarkable manipulation performance and the questions surrounding the demo conditions.
The Astribot S1 made its official public debut at the 2024 World Robot Conference (WRC), held at the Beiren Yichuang International Exhibition Center in Beijing's Economic and Technological Development Zone on August 21, 2024. The 2024 WRC featured 27 humanoid robot manufacturers and over 30 upstream and downstream companies from the humanoid robot supply chain, making it the largest gathering of humanoid robot exhibitors in the conference's history.
At the event, Astribot showcased the S1's full form for the first time, including its wheeled mobile base, dispelling earlier questions about the robot's mobility. The company described the S1 as the "strongest AI robot assistant" and presented it performing a series of challenging, long-sequence, and generalizable tasks at normal speed (as opposed to the accelerated playback speeds sometimes used in robotics demonstrations). Demonstrated tasks included:
CEO Lai Jie stated at the conference that the company aims to "provide several billion people with AI robotic assistants." He acknowledged that the S1's performance at that time reached approximately 55 to 85 percent of human-level competence across different tasks, with the goal of approaching 99.99 percent accuracy over time.
Following the WRC debut, Astribot continued to release demonstration videos and participate in public events. In November 2024, the company demonstrated the S1 making coffee autonomously using Physical Intelligence's pi-zero (pi0) foundation model. In this demonstration, the robot followed high-level instructions such as "make coffee" by decomposing them into smaller actionable steps, including adding water, measuring coffee grounds, and pressing buttons on a coffee machine. The robot's vision system allowed it to identify objects such as mugs and coffee makers even when they were placed in unexpected locations.
For the 2025 Lunar New Year (Year of the Snake) celebrations, Astribot released a video showing its S1 robots preparing for the Spring Festival by cleaning, making tangyuan (glutinous rice balls, a traditional holiday food), and setting off firecrackers. This was part of a broader trend in which Chinese humanoid robot companies participated in the Lunar New Year festivities to demonstrate the practical readiness of their platforms.
The following table summarizes the Astribot S1's key specifications, sourced from the company's official product page and verified third-party reviews:
| Specification | Value |
|---|---|
| Height | 170 cm |
| Weight | Approximately 80 to 90 kg |
| Total Degrees of Freedom (DoF) | 23 |
| Arm DoF | 7 per arm (14 total) |
| Torso DoF | 4 (articulated) |
| Head DoF | 2 |
| Mobile Base DoF | 3 (omnidirectional wheels) |
| End-Effector Maximum Speed | 10 m/s or greater |
| End-Effector Maximum Acceleration | Approximately 100 m/s squared |
| Positioning Repeatability | Plus or minus 0.1 mm |
| Payload per Arm (at full horizontal reach) | 5 kg |
| Combined Bimanual Payload | Up to 10 kg |
| Arm Span | 194 cm |
| Gripper Type | Parallel-jaw (dual-digit) |
| Actuation Type | Cable-driven |
| Battery Type | Lithium-ion |
| Battery Life | 4 to 6 hours active use; up to 10 hours standby |
| Charging Time | Approximately 1.5 hours via docking station |
| Locomotion Speed | 4 km/h |
| Connectivity | Wi-Fi (802.11ac), Gigabit Ethernet, Bluetooth, optional 5G |
The company compares the S1's physical capabilities to those of a normal adult male, noting that the robot matches or exceeds human performance in several metrics. Its positioning repeatability of plus or minus 0.1 mm is substantially better than the typical human hand repeatability of plus or minus 1 to 5 mm. However, the robot's 23 total degrees of freedom are fewer than the approximately 27 degrees of freedom in a human hand alone, reflecting the trade-off inherent in using parallel-jaw grippers rather than articulated fingers.
The S1 carries a comprehensive sensor package that enables autonomous navigation, object recognition, and force-controlled manipulation:
| Sensor Type | Purpose |
|---|---|
| RGB cameras | Visual perception and object recognition |
| Depth cameras (RGB-D) | 3D spatial awareness and distance measurement |
| LiDAR | Spatial mapping and indoor navigation |
| IMU (Inertial Measurement Unit) | Orientation tracking and balance maintenance |
| Force/torque sensors (at wrists) | Contact force measurement for delicate manipulation |
| Tactile pressure sensors (in fingertips) | Grip force feedback and slip detection |
| Ultrasonic proximity sensors | Close-range obstacle detection |
| Microphone array | Voice interaction and sound localization |
| Temperature sensors | Environmental awareness and thermal safety |
The vision system is compatible with pre-trained computer vision encoders including DINOv2, CLIP, and SigLIP, enabling the robot to leverage large-scale visual pre-training for object recognition without requiring task-specific visual training from scratch.
The Astribot S1 uses a cable-driven actuation system inspired by human musculature. According to the company, this design achieves a superior payload-to-weight ratio compared to conventional rigid-link robots, reduces backlash and inertia for smoother movements, and provides compliant motion that yields on contact for safety. The system incorporates what the company describes as a "soft/hard coupling transmission mechanism" with real-time force sensing rather than trajectory estimation, enabling precise control of output force during delicate tasks like handling fragile objects or preparing food.
This cable-driven approach stands in contrast to the direct-drive or harmonic-drive actuators used by many competitors. The compliant nature of cable transmission provides an inherent safety advantage in human-robot interaction scenarios: if a person unexpectedly contacts the robot's arm, the cable system naturally absorbs some of the impact force, reducing the risk of injury.
The company also states that it self-develops key components, including its high-performance motor drive system, to maintain cost advantages and tighter integration between hardware and its AI software stack. This vertical integration strategy mirrors the approach of other successful Chinese robotics firms such as Unitree, which similarly manufactures its own actuators.
Rather than bipedal walking, the S1 uses a three-degree-of-freedom omnidirectional wheeled mobile base. This design choice provides stable, smooth indoor navigation at speeds up to 4 km/h but prevents the robot from handling stairs, uneven terrain, or outdoor environments. The company has acknowledged this trade-off as a deliberate engineering decision to prioritize reliability and manipulation performance for indoor use cases. Other wheeled-base humanoid robots in the market, such as the Galbot G1 and various service robots from Keenon and Pudu Robotics, make similar trade-offs.
The S1 is equipped with parallel-jaw grippers rather than dexterous multi-fingered hands. While this limits the range of grasps the robot can perform compared to five-fingered designs, the parallel-jaw configuration offers high reliability, simpler control, and consistent precision for the household and service tasks the S1 targets. The gripper incorporates tactile pressure sensors in the fingertip surfaces for grip force feedback. Future iterations of the platform are expected to incorporate more advanced dexterous hand designs as the technology matures.
Astribot's core technical philosophy revolves around what it calls "Design for AI" (DFAI), a software-hardware integrated architecture that deeply couples AI capabilities with optimal physical operation. Rather than developing software and hardware independently and integrating them afterward, DFAI co-designs both from the ground up so that the mechanical system is purpose-built for AI-driven control. This approach enables the S1 to learn, plan, and execute tasks in a more tightly integrated manner than would be possible with a generic robot platform running general-purpose AI software.
The S1's primary method of skill acquisition is imitation learning. A human operator demonstrates tasks through a whole-body teleoperation interface using a VR headset and handheld joysticks, and the robot learns to replicate the demonstrated behaviors. This contrasts with approaches that rely heavily on manual programming, reward-function engineering, or simulation-only training.
In July 2025, Stardust Intelligence published a paper on arXiv titled "Towards Human-level Intelligence via Human-like Whole-Body Manipulation" (arXiv:2507.17141), authored by Guang Gao, Jianan Wang, Jinbo Zuo, and colleagues. The paper introduced the Astribot Suite, a comprehensive robot learning framework for whole-body manipulation. The paper described three key contributions:
The DuoCore-WB policy uses RGB-based visual perception compatible with pre-trained vision encoders. It models coordinated whole-body actions using end-effector space control with SO(3) orientation representation, expressed as deltas in the egocentric frame of each end-effector. The framework was evaluated on six representative real-world whole-body tasks:
| Task | Skills Tested |
|---|---|
| Deliver a drink | Long-horizon planning, mobile manipulation, dexterous manipulation with articulated objects |
| Store cat food | Coordinated bimanual manipulation within constrained spaces, dynamic stability with heavy payloads |
| Organize shoes | Whole-body coordination in low-height spaces, synchronous bimanual manipulation |
| Clear a table | Sequential pick-and-place, obstacle avoidance |
| Prepare a meal | Multi-step sequential manipulation with diverse objects |
| Sort laundry | Deformable object manipulation, category recognition |
Across these tasks, the DuoCore-WB policy achieved an average 80 percent success rate, with each policy trained on demonstration data collected via the teleoperation interface. The authors noted that the cohesive integration of embodiment, teleoperation interface, and learning pipeline represents a significant step toward real-world, general-purpose whole-body robotic manipulation.
The company has demonstrated compatibility with external AI foundation models. In November 2024, the S1 was shown making coffee autonomously using Physical Intelligence's pi-zero (pi0) model, a vision-language-action model designed for robotic control. The pi0 model allowed the S1 to accept high-level natural language instructions and decompose them into executable action sequences. Physical Intelligence, a San Francisco-based startup that has raised over $400 million, later open-sourced the pi0 model, enabling broader experimentation.
The company has also indicated that the S1 is undergoing integration and testing with large language models (LLMs) for task planning and natural language understanding. This would allow users to issue verbal commands that the robot interprets and translates into manipulation sequences. As of early 2026, these language-guided capabilities remain in active development.
The Astribot S1 entered limited commercial availability in late 2025, initially in China, with international rollout beginning across the United States, Germany, Japan, South Korea, the United Kingdom, Canada, France, Australia, and India through 2026. Early commercial pricing falls in the range of $96,000 to $150,000, placing the S1 in the premium tier of humanoid robots. Prices are expected to decrease as production scales and component costs decline. For comparison, Unitree's most affordable humanoid model (the R1) starts at approximately $5,900, while the Fourier GR-2 costs approximately $150,000.
The S1 is marketed for several distinct use cases:
Astribot provides a development toolchain that includes comprehensive API access, expert development guidelines, a visual development interface, support for major simulation platforms, and AI deployment guidance. This ecosystem is intended to enable third-party developers and researchers to build applications on the S1 platform, following a model similar to how software development kits are offered for other robotic platforms.
Astribot operates in an increasingly crowded field of Chinese humanoid robot startups, alongside international competitors. The following table provides context on major players in the humanoid robot market as of 2025:
| Company | Country | Notable Product | Approximate Price | 2025 Shipments | Key Differentiator |
|---|---|---|---|---|---|
| AgiBot | China (Shanghai) | AgiBot | Varies | Over 5,100 units | Industrial and commercial focus; leading global market share |
| Unitree Robotics | China (Hangzhou) | G1, H1, R1 | From $5,900 (R1) | Over 4,200 units | Low-cost consumer and research models |
| UBTECH Robotics | China (Shenzhen) | Walker S2 | Enterprise pricing | Approximately 1,000 units | Automotive factory deployment (BYD, Geely) |
| Tesla | United States | Optimus | Projected $25,000 to $30,000 | Approximately 150 units | Integration with Tesla manufacturing ecosystem |
| Figure AI | United States | Figure 02 | Enterprise pricing | Approximately 150 units | Partnership with OpenAI for language-guided control |
| Astribot | China (Shenzhen) | S1 | $96,000 to $150,000 | Limited release | Dexterous manipulation speed and precision |
| Fourier Intelligence | China (Shanghai) | GR-2, GR-3 | Approximately $150,000 (GR-2) | Not disclosed | Healthcare and rehabilitation focus |
| Galbot | China | G1 | Not disclosed | Not disclosed | Hybrid wheeled-bipedal design |
Chinese firms collectively accounted for nearly 90 percent of global humanoid robot shipments in 2025, according to a report cited by Xinhua News Agency. The top Chinese manufacturers, AgiBot and Unitree, each shipped thousands of units, far exceeding the approximately 150 units each shipped by Tesla and Figure AI. This dominance reflects Chinese advantages in supply-chain integration, component cost, and manufacturing scale.
Astribot differentiates itself within this ecosystem through its emphasis on manipulation speed and precision rather than unit volume or price competitiveness. While competitors such as Unitree and AgiBot have focused on scaling shipments at lower price points, and companies like UBTECH have targeted industrial deployment on automotive factory floors, Astribot positions the S1 as a premium platform optimized for dexterous task performance in service and household environments. The 10 m/s end-effector speed and plus-or-minus 0.1 mm positioning repeatability represent the top end of publicly claimed manipulation performance among commercial humanoid robots.
Astribot's development is situated within a broader national push by China to establish leadership in humanoid robotics and embodied AI. Several government initiatives at the central and local levels support this effort:
This policy environment creates favorable conditions for startups like Astribot, providing access to subsidized facilities, government procurement opportunities, and a growing domestic market for humanoid robot deployment. Industry analysts have noted that Chinese humanoid robot companies benefit not only from direct government funding but also from the country's mature electronics supply chain, which provides low-cost sensors, actuators, and computing hardware.
Despite its impressive demonstrations, the Astribot S1 faces several acknowledged limitations:
| Product | Form Factor | Status | Key Features |
|---|---|---|---|
| Astribot S1 | Upper-body humanoid on omnidirectional wheeled base | Commercial (limited availability, late 2025) | 7-DoF arms, 10 m/s end-effector speed, cable-driven actuation, imitation learning, parallel-jaw grippers, 4 to 6 hour battery life |
As of early 2026, the S1 remains the company's sole publicly announced product. No official information about an Astribot S2 or successor model has been confirmed by the company, though the pace of development in the Chinese humanoid robotics sector suggests that updated models are likely under development.