Paper2Agent framework turns research papers into AI agents

A team led by James Zou, Jonathan Pritchard and Jiacheng Miao has published a framework in Nature that turns a research paper into an interactive AI agent in about 45 minutes for roughly $14 on a personal laptop.

The framework, called Paper2Agent, uses the Model Context Protocol to expose a paper's code, data and workflows as callable tools that a chat agent such as Claude Code can invoke through natural language. The abstract frames the goal as transforming research output from "passive artifacts" into "active systems that can accelerate downstream use, adoption, and discovery."

In the headline case study, Paper2Agent built 22 MCP tools around AlphaGenome. The resulting agent scored 98.7% on 15 tutorial queries and, on 15 novel queries, reached "100% accuracy, faithfully producing the expected numerical outputs." Smaller agents for Scanpy and TISSUE, covering single-cell and spatial transcriptomics analyses, were built the same way.

Across a wider benchmark drawn from 100 computational biology papers, the authors report 91.2 ± 1.6% accuracy on tutorial-style questions. Five researchers we track pushed the paper into their feeds the same week.

The authors close on their own framing: "By turning static papers into dynamic, interactive AI agents, Paper2Agent introduces a new paradigm for knowledge dissemination."