LiveCodeBench
Continuously refreshed coding benchmark built from new contest problems to limit training-data contamination.
Reference by Researchers at UC Berkeley, MIT and Cornell
IndexedMIT★ 964
About
How well a model generates, repairs and reasons about code on contest problems published after its training cutoff.
Use cases: llm-evaluation, code-generation
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
About the benchmark
- Benchmark site
- livecodebench.github.io/
- Current version
- release_v6 (1,055 problems)github.com
- Maintainer
- Researchers at UC Berkeley, MIT and Cornellarxiv.org
- Launched
- 2024-03arxiv.org
- Leaderboard
- livecodebench.github.io/leaderboard.html
We link to the official leaderboard instead of copying scores. Benchmarks get gamed and saturate; we never use them to rank listings.
Signals
- Stars
- 964
- Forks
- 208
- Open issues
- 39
- Last commit
- Jul 16, 2025 (1 year ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- release_v6 (1,055 problems)
- License
- MIT
- Repository
- github.com/LiveCodeBench/LiveCodeBench
- Website
- livecodebench.github.io/
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
Cite this page
LiveCodeBench benchmark. Agent Mart. https://agentmart.dev/reference/livecodebench (last updated 2026-10-24)