MedAgentsBench: Benchmarking Thinking Models and Agent Frameworks for Complex Medical Reasoning

TL;DR

MedAgentsBench is a new benchmark for evaluating medical reasoning in AI models, emphasizing complex multi-step clinical tasks to better differentiate advanced models and analyze their performance, cost, and inference time.

Contribution

We introduce MedAgentsBench, a comprehensive benchmark addressing limitations of existing medical QA evaluations, and provide systematic analysis of model performance on complex reasoning tasks.

Findings

DeepSeek R1 and OpenAI o3 excel in complex medical reasoning

Search-based agent methods show favorable performance-to-cost ratios

Significant performance gaps exist between model families on complex questions

Abstract

Large Language Models (LLMs) have shown impressive performance on existing medical question-answering benchmarks. This high performance makes it increasingly difficult to meaningfully evaluate and differentiate advanced methods. We present MedAgentsBench, a benchmark that focuses on challenging medical questions requiring multi-step clinical reasoning, diagnosis formulation, and treatment planning-scenarios where current models still struggle despite their strong performance on standard tests. Drawing from seven established medical datasets, our benchmark addresses three key limitations in existing evaluations: (1) the prevalence of straightforward questions where even base models achieve high performance, (2) inconsistent sampling and evaluation protocols across studies, and (3) lack of systematic analysis of the interplay between performance, cost, and inference time. Through experiments with various base models and reasoning methods, we demonstrate that the latest thinking models, DeepSeek R1 and OpenAI o3, exhibit exceptional performance in complex medical reasoning tasks. Additionally, advanced search-based agent methods offer promising performance-to-cost ratios compared to traditional approaches. Our analysis reveals substantial performance gaps between model families on complex questions and identifies optimal model selections for different computational constraints. Our benchmark and evaluation framework are publicly available at https://github.com/gersteinlab/medagents-benchmark.