Archive — CN — technode
Alibaba’s Qwen releases open-source model for autonomous driving
Alibaba's Qwen team launched Qwen-Drive-1.0-4B, an open-source autonomous driving model developed with Huazhong University. It utilizes vision-language technology for driving-scene understanding and vehicle motion planning, offering two planning versions and released under Apache 2.0 license.
Key points
- Qwen-Drive-1.0-4B is an open-source model for autonomous driving.
- Developed in collaboration with Huazhong University of Science and Technology.
- Utilizes Qwen3.5-4B as its vision-language foundation for enhanced driving-scene understanding.
- Includes components for 3D perception and trajectory generation.
- Offers two versions of planning: one trained to imitate examples, another optimized via reinforcement learning.
- Model and components released under Apache 2.0 license, allowing public access.
Why it matters
- This release aligns with the 2025 IT trend of open-source software driving innovation in AI and autonomous systems.
- It reflects the increasing collaboration between tech companies and academic institutions in pushing the boundaries of AI research.
- The trend of integrating AI with mobility solutions is prevalent as cities aim for smarter, safer transportation systems.
Alibaba’s Qwen-Drive-1.0-4B represents a significant step toward advancing open-source contributions in autonomous driving, potentially accelerating innovation in AI-driven mobility solutions.