Paper2Agent turns research papers into interactive AI agents

September 16, 2026 at 08:47 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Stanford researchers unveiled Paper2Agent, an AI tool that converts a research paper into a bespoke agent capable of answering questions and applying the paper's methods to new data. The system built an agent for the AlphaGenome paper in about 45 minutes at a computing cost of $14. The agent answered genetics questions with near-perfect accuracy and outperformed other biomedical AI agents.

Key Facts

Paper2Agent converts a research paper into an AI agent that serves as a virtual corresponding author, answering questions and applying the paper's methods to new data.
The system built an agent for the AlphaGenome paper in about 45 minutes at a computing cost of $14.
The AlphaGenome agent answered genetics questions with near-perfect accuracy and outperformed other top biomedical AI agents, including Biomni.
The agent identified a different causal gene than the one pinpointed in the original AlphaGenome paper when asked about a genetic variant linked to bad cholesterol.

Paper2Agent System

Paper2Agent accesses a paper's main text, code, data sets and other elements, depositing the information onto a digital platform called an MCP server. A team of AI agents autonomously builds tools that apply the paper's methods to fresh data and places those tools on the server. Scientists connect to the server using a large language model of their choice, creating a paper-specific agent that can be interacted with in plain, conversational language. James Zou, a computer scientist at Stanford University and co-author of the paper, says the tool can help reimagine what knowledge looks like in the future.

AlphaGenome Test

Zou and his team tested the technology on the paper that introduced AlphaGenome, an AI model that predicts the properties of DNA sequences. Paper2Agent autonomously created an agent for the AlphaGenome paper in about 45 minutes, with required computing power costing $14. The agent answered genetics questions with near-perfect accuracy and bested the scores of other top biomedical AI agents given access to the same paper. Among the outscored agents was Biomni, a tool developed by academic researchers that draws from dozens of databases but scored far lower than the AlphaGenome agent. Zou says the paper agent's success comes from its mastery of AlphaGenome's tools and abilities.

Genetic Variant Analysis

The team asked the agent to determine why a single change to a DNA letter in a genetic sequence is associated with bad cholesterol. The agent identified a gene that could explain the link, but it was a different causal gene than the one pinpointed in the original AlphaGenome paper. Zou says AlphaGenome's data on genetic variants support both hypotheses.