Global Robot LLM Market to Exceed $100B by 2028, NVIDIA’s WFM Platform to Drive Growth

Estimated reading time: 2 minutes

TrendForce’s latest investigations report that as humanoid robots move toward highly integrated systems and transition from industrial applications to home environments, AI model training will become increasingly critical to meet the growing demands for backend understanding and interaction capabilities. The global LLM market for robotics, including AI training and AIGC solutions, is expected to surpass US$100 billion by 2028, with a CAGR of 48.2% from 2025 to 2028.

At CES 2025, NVIDIA introduced its Cosmos platform, which integrates the World Foundation Model (WFM). This platform is expected to become a key driver of market growth.

The Cosmos platform—closely tied to robot development—leverages real-world data and AI-generated data to build digital twin environments for training. This approach offers advantages such as cost reduction, simplified processes, customized training, and more realistic interaction simulations.

NVIDIA aims to accelerate the development of general-purpose robotics technologies with Cosmos while recognizing the evolution of AI from generative AI to agentic AI and, eventually, physical AI. This progression is crucial for autonomous vehicles and robots to advance training processes and achieve broader commercial deployment.

One of the key challenges in robot training is the sim-to-real gap—the discrepancy between virtual training environments and real-world conditions. Deep learning and neural networks primarily rely on simulated data, which can lead to inefficiencies in real-world performance due to a lack of alignment with physical properties, sensor data, and actuator dynamics.

NVIDIA introduced the NaVILA model in December 2024 to bridge this gap, enabling robots to understand natural language commands without pre-training or mapping and navigate complex terrains using visual systems and LiDAR technology.

Currently, AI training and AIGC solutions account for 10% of humanoid robot costs, representing the largest cost share on the software side. Other cost contributors include operating systems (4%) and autonomous navigation (3%).

Several key players in NVIDIA's ecosystem are driving advancements in robotic simulation technologies. Companies such as Siemens, Vention, and Hexagon specialize in robot simulation, while Rockwell Automation, PTC, and Cadence focus on computer-aided engineering (CAE) and reality capture. Additionally, MetAI, a Taiwanese startup, contributes innovative solutions to the ecosystem.

TrendForce emphasizes that NVIDIA’s Cosmos platform is poised to address the training gap challenges, highlighting the importance of AI large model simulations. This development marks a critical step toward commercializing humanoid robots and integrating them into everyday life.

Leading humanoid robot developers, including Figure AI and 1X, are expected to be among the early adopters of the Cosmos platform, setting the stage for a new era of AI-driven robotic innovation.