XPeng MONA M03 2026: 10K Orders in 37min

🔑 Enhanced Key Takeaways

• •The MONA M03 series utilizes a 'Vision-only' autonomous driving architecture, moving away from LiDAR to reduce costs and achieve the aggressive sub-160,000 RMB price point.
• •The 750TOPS compute power is driven by dual NVIDIA Orin-X chips, which are integrated into the Max and Ultra SE trims to support the upcoming VLA (Vision-Language-Action) model deployment.
• •XPeng's strategy for the MONA sub-brand focuses on 'AI-defined mobility' for the mass market, leveraging the company's proprietary end-to-end large model technology previously reserved for flagship models like the X9 and G9.

🛠️ Technical Deep Dive

• •Architecture: Utilizes a centralized E/E architecture that supports OTA updates for the entire vehicle, specifically optimized for the VLA (Vision-Language-Action) model.
• •VLA Implementation: The 2nd-gen VLA model uses a transformer-based architecture that maps visual inputs directly to control commands, bypassing traditional rule-based coding for complex urban navigation.
• •Battery/Range: Features a high-energy-density LFP battery pack with a drag coefficient of 0.194Cd, contributing to the 540km/640km CLTC range ratings.
• •Compute: The Max/Ultra SE trims leverage dual NVIDIA Orin-X SoCs, providing 750 TOPS of AI performance, allowing for real-time processing of high-resolution camera feeds without LiDAR input.

🔮 Future ImplicationsAI analysis grounded in cited sources

⏳ Timeline