WeRide and Lenovo are expanding their collaboration to deploy 200,000 autonomous vehicles worldwide over five years starting in 2026, including Robotaxis, as part of a major push to commercialize Level 4 autonomous driving.

WeRide already operates across more than 40 cities in 12 countries, supported by its WeRide One platform and commercial model, forming a global footprint for the rollout.

The agreement was signed at Auto China 2026, with top executives from both firms in attendance, signaling a strategic alliance aimed at mass deployment.

Key challenges identified include maintaining consistent AV performance across diverse urban environments and regulatory regimes, while balancing high-performance computing with cost efficiency and reliability.

The collaboration focuses on moving from pilots to scalable commercial deployment, balancing computing power, system reliability, and cost, and advancing Physical AI applications in real-world mobility.

The platform is designed to meet global safety and regulatory requirements with auto-grade hardware and fully redundant software, enabling scalable Level 4 deployment while reducing autonomous driving suite costs and total cost of ownership.

The news was issued as a business update and RTT News noted; for comments, [email protected] can be contacted.

HPC 3.0 further reduces costs and enables scalable deployment, while leveraging NVIDIA DRIVE AGX Thor for robust computing power.

Overall, HPC 3.0 aims to cut the autonomous driving suite cost by about half and lower TCO by roughly 84% over its lifecycle versus HPC 2.0.

The collaboration stresses cross-cloud and vehicle integration to accelerate real-world mobility deployment, including autonomous minibuses and sanitation vehicles as part of future plans.

WeRide emphasizes that autonomous driving represents a pivotal phase for commercial deployment where cost efficiency and scalable rollout are increasingly decisive for success.

The joint HPC 3.0 platform, deployed on the WeRide Robotaxi GXR and built around Lenovo AD1 and NVIDIA DRIVE AGX Thor, delivers over 2,000 TOPS and reduces autonomous-driving suite costs and total cost of ownership significantly compared with earlier generations.