Alibaba releases Qwen-Drive: Autonomous driving model preserving base VLM
aigclink · x · 2026-09-02
Alibaba released Qwen-Drive-1.0, an autonomous driving model that preserves the pre-trained VLM architecture, differing from end-to-end specialized models by plugging in a 3D perception head and a planning expert.
Key Features:
- BEV Head: Enables 3D object detection, semantic occupancy prediction, and map segmentation.
- Planning Expert: Uses flow matching to generate a 5-second trajectory (50 waypoints).
- Performance: Achieves a NAVSIM score of 90.7, comparable to SOTA planners.
Related event: Alibaba Releases Qwen-Drive-1.0: Unified Perception and Planning VLM(4 posts)→
More from Models
- Jev sparks debate: training your own classifier won't beat Voyage's years of reranker tuning — zainhas · 2026-09-21
- Freebuff's $8/Month Ad-Supported Coding Sub Promises 15 Hours of DeepSeek Daily — gaganghotra_ · 2026-09-21
- The Illustrated Recurrence: From Amari-Hopfield Nets to GPT-6 Astra — gklambauer · 2026-09-21
- Unverified claim: Grok 4.7 launching today with 500K context, multimodal, 4.6-level pricing — realsohamparekh · 2026-09-21
- Dev warns: if token subsidies end, 24/7 agent use gets priced out alongside local — BLUECOW009 · 2026-09-21
- Simulated test: GPT-6 Astra pushed a virtual person off a ledge in multiple trials, rivals didn't — paul_cal · 2026-09-21