Beijing (Gasgoo)- Global mobility tech company ECARX has successfully deployed and deeply adapted the DeepSeek-R1 AI large model to its ECARX AutoGPT in-vehicle AI system, according to the company's announcement on Feb. 10.
This major upgrade expands AutoGPT's edge computing capabilities and enhances its inference, decision-making, and specialized expertise for vertical domains.
ECARX's AutoGPT is an exclusive in-car large model built on a general-purpose language model, tailored to mobility scenarios. It integrates four core capabilities: AutoAgent AI, AutoFlow AI task execution, AutoScene AI scenario engine, and AutoEco AI service ecosystem. The system can autonomously perform complex tasks, understand natural language, and support multimodal perception, offering a highly personalized and intelligent travel experience.
The DeepSeek-R1 integration significantly boosts ECARX AutoGPT's AI scenario engine, particularly in complex cross-domain decision-making scenarios. DeepSeek-R1, using reinforcement learning methods, improves reasoning, generalization, and decision recommendation abilities. This on-device model processing reduces dependency on cloud resources, ensuring data privacy, lower latency, and a more secure AI experience, making it ideal for global deployment with local data handling and reduced certification costs.
