Pittsburgh-based Skild AI, the robotics software company building what it calls the industry's first unified robotics foundation model, has closed a $1.4 billion Series C led by SoftBank Group. The round values Skild at more than $14 billion, more than tripling the $4.5 billion valuation set by its $135 million Series B just seven months earlier and ranking the company among the most valuable robotics startups in the world.
Investor syndicate
NVIDIA's venture arm NVentures, Bezos Expeditions, Macquarie Capital, Disruptive and 1789 Capital all participated alongside SoftBank. The round also drew a long roster of strategic checks from Samsung, LG Technology Ventures, Schneider Electric, CommonSpirit Health and Salesforce Ventures, signaling that industrial conglomerates, healthcare systems and enterprise software vendors all want a seat at the table as humanoid and mobile manipulator fleets begin to scale.
The Skild Brain
Skild's pitch is that the Skild Brain is body-agnostic: a single neural network trained on internet-scale human video, lab data and physics-based simulation that can drive quadrupeds, humanoids, tabletop arms and AMRs without per-robot retraining. The company has grown from zero to roughly $30 million in revenue in months as it has begun licensing the model to OEMs and integrators. The funding will go toward expanding the model, growing the data engine and supporting on-robot deployments at customer sites.
Context: physical AI's mega-round era
Skild's raise comes as physical AI funding hits records. Apptronik's $520 million extension, Mind Robotics' $400 million, and Pudu's $150 million all closed in the same quarter, while Project Prometheus' $10B round reset the upper bound for the sector. Investors now openly describe robotics foundation models as the next NVIDIA-scale platform opportunity.
What Skild has to prove
The challenge for Skild will be turning its omni-bodied claim into commercial wins against vertically integrated humanoid OEMs that train their own models on their own hardware. Customers will look for evidence that one model can outperform purpose-built stacks on safety, dexterity and uptime at the price points industrial buyers expect. Management said the Series C extends the company's runway through at least 2028 and funds the first wave of large-volume customer deployments.
Reporting based on coverage from TechCrunch, The Robot Report, BusinessWire, Crunchbase News and SiliconANGLE.