Astribot, completed a Series B financing round this week. Chinese reports put cumulative financing across three rounds in three months above RMB 1 billion, with valuation now above RMB 10 billion, or roughly .4 billion. China Renaissance acted as exclusive financial adviser.

Astribot is pairing new capital with a lower-cost robot, a cable-driven manipulation architecture, and a self-developed stack spanning model layer, embodied operating system, and robot body. The company is trying to lower the cost of putting manipulation hardware into more real-world settings.

That price sits far below the company's earlier S-series robot. Astribot's S1, its higher-end flagship manipulation platform, is priced at roughly RMB 690,000.

T1 lowers the entry point for researchers, developers, education users, commercial-service operators, entertainment venues, and industrial pilots.

The technical route

Astribot’s route is different from many humanoid developers.

The company describes its robots as rope-driven systems, using a tendon-like cable transmission approach. Its stated design philosophy, which it calls Design for AI, treats the robot body as a surface built for AI training rather than a conventional hardware platform with software added afterward.

The argument is that useful embodied AI depends on dexterity, force control, contact recovery, and physical interaction data. Astribot’s pitch is that those capabilities are easier to build when the robot body, operating system, and model layer are designed together.

In the company's own framing, the stack is an AI model layer, an embodied operating system, and a rope-driven robot body.

S1 established the company’s public profile through manipulation demos: folding clothes, pouring, sorting objects, tool use, and high-speed handling.

Locomotion demos generate attention, but manipulation is one of the harder bottlenecks in embodied AI. Useful robots need to handle objects, recover from failed grasps, operate in messy environments, and generate dense physical-interaction data. For most commercial settings, walking is not enough. The robot has to do something with its hands.

Astribot has also published model work around that direction. Its Lumo-1 research describes a vision-language-action model for robot execution, with training tied in part to trajectories collected from Astribot S1.

The company lineage

Stardust Intelligence was founded in Shenzhen in 2022 by Lai Jie and a founding team with roots in Tencent Robotics X.

Lai previously led Baidu's Xiaodu Robot team and was reportedly employee number one at Tencent Robotics X, where he led development of Ollie, a wheel-legged robot known for its tail-assisted backflip.

The round

Investors in the Series B include Liangxi Science and Technology Innovation Industry Phase II Mother Fund, managed by Bohua Capital, Yangzhou Longtou Core Chip, Zhongbo Juli, ThunderSoft, Kede Education, and existing backer Guoke Investment, the investment arm of the Chinese Academy of Sciences.

ThunderSoft is a Shenzhen-listed embedded OS and automotive AI company. Alongside the funding participation, Astribot and ThunderSoft are reported to be starting a 1,000-unit embodied-robot deployment cooperation covering industrial manufacturing and commercial-service scenarios, with joint work on robot operating systems, VR interaction, and robot-intelligence applications.

If the cooperation materializes, ThunderSoft can provide more than capital, software infrastructure, deployment surface, and potential channel partner around the same embodied-robot stack Astribot is trying to scale.

The 1,000-unit cooperation is reported, but no named customer, delivery confirmation, operating site, or revenue detail has been published.

The China context

Astribot is not moving in isolation.

China's embodied AI market is producing a fast cluster of high-valuation robotics companies and component suppliers. Unitree has moved toward a Shanghai STAR Market IPO. LinkerBot has been reported to be aiming for a billion valuation target for robotic hands. AgiLink, a robotic-hand company linked to AgiBot, reportedly crossed a billion valuation after multiple rounds since its January spinout.

The common thread is deeper than purely humanoids, It is hardware cost, component scale, dexterous manipulation, data collection, and public-market preparation. China’s robotics market is starting to see increased platform and supply-chain competition.

Have a robotics update Korthos should review? Send news, deployments, product releases, funding rounds, research, or media to tips@korthos.xyz or reach out on X at @agkorthos.