Fysics

Summary

About

About Fysics

Fysics develops simulation software for modelling how robots and physical systems interact with rigid bodies, soft materials, fluids and particles. Its stack includes the Fysics engine, MoziSim training environment, OmniFysics model, FysicsWorld and FysicsEval benchmarks.

Fysiverse reconstructs geometry and material properties, computes motion, contact and deformation through the physics engine and conditions generated visual outputs on those states. The Shanghai company completed three Pre-A closings in 2026, with cumulative financing described as several hundred million renminbi. Commercial milestones remain concentrated in technical releases rather than named deployments.

PRICE
Range 1Y
Return
High
Low
Updated

Articles

Funding

Investors

  • Changlei Ventures - Participant - VC - 1 funding event - first seen 2026-05-25
  • Changshi Capital - Participant - VC - 1 funding event - first seen 2026-05-25
  • Eastern Bell Capital - Participant - VC - 1 funding event - first seen 2026-05-25
  • Pangu Ventures - Participant - VC - 1 funding event - first seen 2026-05-25
  • Silicon Harbour Capital - Participant - VC - 1 funding event - first seen 2026-05-25
  • TEDA Fund - Participant - VC - 1 funding event - first seen 2026-05-25
  • Yunqi Partners - Lead - VC - 1 funding event - first seen 2026-05-25
  • Zhidao Capital - Lead - VC - 1 funding event - first seen 2026-05-25
  • Zhongying Ventures - Participant - VC - 1 funding event - first seen 2026-05-25

Team

  • Lihua Zhang - Founder - Founder and physics-engine technical leader.
  • Dingkang Yang - Chief Technology Officer - Chief technology officer responsible for the Physical-AI stack.

Timeline

Events

  • 2026-06-25 - Product Launch - Fysics introduces Fysiverse physical world model

    The pipeline reconstructs geometry and material properties from observations before running physical evolution through the Fysics engine. Motion, contact, occlusion and deformation outputs condition the generative renderer rather than relying on trajectory prompts alone. The architecture is intended for robot policy training, parameter estimation and physical-outcome prediction.

    Source
  • 2026-05-25 - Funding - Fysics closes Pre-A3 financing

    Zhidao Capital and Yunqi Partners co-led the round, with Eastern Bell Capital and Silicon Harbour Capital continuing as investors. TEDA Fund, Changlei Ventures, Changshi Capital, Pangu Ventures and Zhongying Ventures also participated. The three disclosed Pre-A closings reached cumulative financing of several hundred million renminbi.

    Source
  • 2026-03-27 - Product Launch - Fysics launches differentiable physics engine and Physical-AI stack

    The engine supports coupled rigid-body, soft-body and fluid simulation with differentiable optimisation across the simulation and training loop. MoziSim covers scene construction, robot and sensor simulation, training and sim-to-real workflows for humanoids, quadrupeds and robot arms. OmniFysics, FysicsWorld and FysicsEval add model and benchmark layers around the simulation stack.

    Source

Context

Comparable profiles

  • Space Data

    stack adjacency - Simulation · Physics AI Simulation

  • Simularge

    stack adjacency - Simulation And Test Infrastructure · Physics AI Simulation

  • Treble Technologies

    stack adjacency - Simulation · Physics AI Simulation

  • RoboDK

    stack adjacency - Simulation · Simulation And Test Infrastructure

  • Simumatik

    stack adjacency - Simulation · Simulation And Test Infrastructure

  • Visual Components

    stack adjacency - Simulation · Simulation And Test Infrastructure

  • bestat

    stack adjacency - Simulation

  • Palatial

    stack adjacency - Simulation

  • Fern

    stack adjacency - Simulation And Test Infrastructure

  • MORAI

    stack adjacency - Simulation And Test Infrastructure

  • dSPACE

    stack adjacency - Simulation And Test Infrastructure

  • ReSim

    stack adjacency - Simulation And Test Infrastructure

Context tags

  • Simulation - Stack Layer - Differentiable physics engine and simulation-training stack: https://m.huanqiu.com/article/4QvFMULQf4n
  • Simulation and Test Infrastructure - Stack Layer - Simulation and test infrastructure for robot and Physical-AI development: https://news.pedaily.cn/202605/564303.shtml
  • Physics AI simulation - Capability - Physics-AI simulation and physically constrained world modelling: https://tech.chinadaily.com.cn/a/202606/25/WS6a3cd0baa310d709c2fba14e.html