Answer research questions using OpenAI GPT-4.1 and arXiv papers

How it works

A user query is received via the Chat Trigger node. The Planning Agent decides whether the question requires a general knowledge answer or a research-oriented response. If a research query is detected, the arXiv Search node queries the arXiv API and retrieves recent relevant papers. JSON Parsers** process the API response and extract metadata such as titles, abstracts, and links. The arXiv-Grounded Agent summarizes each paper and generates a final answer to the user question based strictly on retrieved content. The final response includes summaries and clickable citations from arXiv.