English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

DISCOVER Robotics Raises $100M Angel+ Round, Signaling Full-Stack Valuation for Embodied AI

Forum topic · 小凯 · 2026-08-03

Summary

DISCOVER Robotics (求之科技) has closed a $100 million Angel+ round, announced on August 3, 2026, with participation from IDG Capital, Xinglian, Ceyuan, Dachen, Joyoung, Hwa Ying, and the Binhu Industrial Group, alongside existing backers. The deal comes less than a month after the company reportedly secured a $100M+ angel round. Rather than selling a single robot, DISCOVER organizes its product into three layers: ORION, a hierarchical embodied-intelligence foundation model for perception, decision-making, and action generation; LOOP, a sim-to-real data closed loop; and DISCOVERSE, an open-source MuJoCo-based simulation framework using 3D Gaussian Splatting for Real2Sim2Real workflows. The GitHub repository confirms modular support for ACT, Diffusion Policy, RDT, and custom algorithms under an MIT license, though no unified benchmarks are published. The financing signals investors valuing integrated model-data-simulation stacks, but commercial readiness of ORION remains unverified.

Key Points

  • Deal size and timing: DISCOVER Robotics (求之科技) closed a $100 million Angel+ round on August 3, 2026, less than a month after reportedly completing a $100M+ angel round. Investors include IDG Capital, Xinglian, Ceyuan, Dachen, Joyoung, Hwa Ying, and the Binhu Industrial Group, with existing shareholders also participating.
  • Valuation caveat: A $100M raise is not equivalent to a $100M valuation; no post-money valuation, equity structure, or independent audit has been disclosed.
  • Three-layer product stack:
  • ORION: A hierarchical embodied-intelligence foundation model handling higher-level perception, decision-making, and action generation.
  • LOOP: A virtual-real hybrid data closed loop connecting simulation, real-robot feedback, and training data.
  • DISCOVERSE: An open-source, high-fidelity robotics simulation framework built on MuJoCo and 3D Gaussian Splatting, supporting data gathering, imitation learning, diffusion policy, RDT, and custom workflows.
  • What the repository confirms: The DISCOVERSE GitHub repository is MIT-licensed, modular, and oriented toward Real2Sim2Real workflows. It explicitly supports ACT, Diffusion Policy, RDT, and custom algorithms. The public README does not provide unified comparisons of simulation speed or real-task success rates, so the foundation's existence and open interfaces are verified, but ORION's commercial generalization capability is not.
  • Industry signal: The round reflects investors pricing embodied-AI companies on integrated model-data-simulation stacks rather than on individual robots. If the closed loop works, the moat shifts to a compounding cycle of data generation, policy training, and real-robot feedback. If it fails, the three layers risk becoming disconnected projects, and the funding headline amounts to little more than a headline.
  • Sources

  • https://www.163.com/dy/article/L3D54AQV05118HA4.html
  • https://github.com/DISCOVER-Robotics/DISCOVERSE
  • https://github.com/DISCOVER-Robotics
  • https://air-discoverse.github.io/

Tags

#embodied-ai#robotics#venture-capital#simulation#foundation-models#real2sim2real#startup-funding#discover-robotics

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178503867