Since its founding in 2021, DexForce has focused on physical AI and world models, pursuing an integrated approach that combines world modeling, physics simulation, and real-world robot deployment. Its proprietary DexWorldModel is described as a core engine that models three-dimensional physical laws with dynamic latent-space modeling, unified coordinate alignment, and precise spatiotemporal matching. The company says this enables improved spatial perception and physical modeling to support tasks such as sorting, assembly, and dual-arm manipulation. DexForce claims to have outpaced larger peers on several technological milestones and recently topped the WorldArena benchmark for embodied world models. Financially, it generated ¥100 million in revenue in the first half of the year and is on track to surpass ¥300 million for the full year. Post-Series B, the company plans to invest in algorithm iteration, physics engine upgrades, data infrastructure, and further development and deployment of humanoid robot capabilities.

Total raised

$157M

Funding rounds

5

Latest round

Series B

Latest activity

Jun 2026