Tag
UTP-Bench is a new benchmark for uncertainty-aware travel planning that evaluates LLMs on robustness against delays and crowd variability using real-world data from India.
Google is enhancing its AI Mode with new features to track flight prices and book hotels, positioning it as an AI travel agent for trip planning and booking.
Google enhances AI Mode in Search with new travel features including flight price tracking, points/miles rates viewing, and in-app hotel booking for a seamless trip planning experience.
Introduces TREK, a benchmark for evaluating LLM agents on complex travel planning tasks with deterministic scoring, covering 800 multi-constraint tasks over a synthetic knowledge base. The benchmark reveals that even strong agents struggle with unstated user needs.
Sam Altman shares a personal demonstration of ChatGPT's ability to handle a complex, multi-step request involving travel planning, site creation, and email drafting, highlighting the model's impressive capabilities.
This paper proposes AI Tour Meeting, a group travel planning framework that uses multiple LLM-based agents with distinct personas to collaboratively find itineraries through natural language discussion.
This paper evaluates hand-off compression in a two-agent LLM relay for travel planning, comparing methods like JSON extraction and narrative summarization, finding that structured representations preserve constraints better.
A developer created a 12306 ticket query MCP server based on MCP, offering a simple API for AI assistants to directly query ticket information, supporting functions such as ticket search, stopover query, and transfer query.
OpenAI has released a new voice model, GPT-Live, capable of natural multi-task conversations and real-time execution of complex planning, such as arranging a one-day trip from Tokyo to Dubai to Hawaii.