Travel itinerary planning represents a pivotal challenge in artificial intelligence, combining real-world complexity with the need for reliable, user-centric decision-making. As a ubiquitous yet intricate task, it demands seamless integration of spatial, temporal, and financial constraints while aligning with diverse user preferences such as budget limits, dining habits, and activity priorities. Despite advancements in AI-driven planning systems, existing solutions often struggle to balance long-horizon reasoning, multi-modal data fusion, and dynamic constraint satisfaction—highlighting a critical gap between academic research and practical deployment.

The IJCAI-25 Travel Planning Challenge (TPC@IJCAI25), co-organized by Nanjing University, Huawei, and JSAI, aims to bridge this gap by fostering innovations in AI agents capable of generating robust, personalized travel plans. Participants will tackle the task of synthesizing detailed itineraries—spanning transportation, accommodation, activities, and cost management—based on user queries with heterogeneous requirements. This competition intersects cutting-edge domains such as Natural Language Processing, LLM Reasoning and Planning, Constraint Optimization, and Neuro-Symbolic Learning, offering a unique platform to explore the synergy of generative AI, recommendation systems, and planning technologies.