LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming?

TL;DR

This paper evaluates the limitations of current large language models in competitive programming by introducing a new benchmark and analysis, revealing they lag behind human experts especially in complex reasoning tasks.

Contribution

The paper presents LiveCodeBench Pro, a new benchmark with expert annotations, and provides detailed diagnostics of LLMs' performance gaps in competitive programming.

Findings

LLMs achieve only 53% pass@1 on medium problems without tools

LLMs score 0% on hard problems, where humans excel

Performance driven by implementation skills, not reasoning

Abstract

Recent reports claim that large language models (LLMs) now outperform elite humans in competitive programming. Drawing on knowledge from a group of medalists in international algorithmic contests, we revisit this claim, examining how LLMs differ from human experts and where limitations still remain. We introduce LiveCodeBench Pro, a benchmark composed of problems from Codeforces, ICPC, and IOI that are continuously updated to reduce the likelihood of data contamination. A team of Olympiad medalists annotates every problem for algorithmic categories and conducts a line-by-line analysis of failed model-generated submissions. Using this new data and benchmark, we find that frontier models still have significant limitations: without external tools, the best model achieves only 53% pass@1 on medium-difficulty problems and 0% on hard problems, domains where expert humans still excel. We also find that LLMs succeed at implementation-heavy problems but struggle with nuanced algorithmic reasoning and complex case analysis, often generating confidently incorrect justifications. High performance appears largely driven by implementation precision and tool augmentation, not superior reasoning. LiveCodeBench Pro thus highlights the significant gap to human grandmaster levels, while offering fine-grained diagnostics to steer future improvements in code-centric LLM reasoning.