Benchmark
LiveCodeBench
Compare coding models on dated contest problems and code-related reasoning tasks.
Published leaderboard
Explore scores and evaluation settings on the publisher’s leaderboard.
Open leaderboardCoverage
Problems collected over time from LeetCode, AtCoder, and Codeforces. The project covers code generation, self-repair, code execution, and test-output prediction.
Methodology
Select a dated problem window and evaluation scenario, then score generated solutions or predicted outputs against the corresponding tests.