On November 11, 2026, autonomous driving software developer TIER IV announced the release of multiple end-to-end AI models designed to work with Autoware. The models provide developers with flexible configuration options ranging from camera-only systems to setups incorporating LiDAR, radar, and HD maps.
New AI Model Portfolio
TIER IV developed and open-sourced multiple end-to-end autonomous driving AI models to support varying system requirements, including map information and sensor configurations. The models offer options from camera-only setups to configurations incorporating LiDAR, radar, and high-definition maps. Developers can use Autoware to optimize these models and integrate them into vehicle systems.
According to TIER IV founder and CEO Shinpei Kato, the initiative expands choices for automakers and component suppliers to internalize AI technology according to their applications and performance requirements.
Overview of Released Models
The release includes NVIDIA Alpamayo 2 Edge, a model integrating vision-language-action capabilities with Autoware that takes camera images as input and uses reasoning to generate driving trajectories. AutoE2E, led by the Autoware Foundation and TIER IV, features a Fast-Slow architecture and runs on computing platforms beyond NVIDIA hardware. METEOR is a lightweight multitask model built on a convolutional neural network architecture for privately owned passenger vehicles.
Additional models include OnePlanner, which integrates perception with motion planning, and LatentDrive, which uses a world model approach with V-JEPA for trajectory generation. TIER IV stated that it will continue improving the performance of the AI models and optimizing them for in-vehicle environments.
Key facts and where they come from
- TIER IV released multiple end-to-end autonomous driving AI models supporting different system requirements.
announced the development and open-source release of multiple end-to-end (E2E) autonomous driving AI models that support different system requirements, including map information and sensor configurations.
- NVIDIA Alpamayo 2 Edge features 2 billion parameters and generates driving trajectories from camera images.
With 2 billion parameters, compared with 34 billion in Alpamayo 2 Super, it is substantially smaller and offers promise for practical deployment in vehicles.
- METEOR uses 16 million parameters and supports operation across semiconductor platforms including NVIDIA and Renesas.
With 16 million parameters, it is even lighter than AutoE2E and supports operation across different semiconductor platforms, including NVIDIA and Renesas.
