LetTokenBurn
Sign in
Verified100%Harbor nativeCodingAgentic / Tool Use

LiveCodeBench

Live Code Bench

LiveCodeBench: 持续更新竞赛编程题

PaperRepoDatasetLeaderboardHomepage

How to run

ltb · livecodebench
$ npm i -g lettokenburn-cli
$ ltb bench run livecodebench/livecodebench@1 --benchmark livecodebench -a claude-code -m anthropic/claude-opus-4-8
$ ltb bench push

Harbor-native: you can also call Harbor directly.

harbor
$ harbor run -d livecodebench/livecodebench@1 -a claude-code -m anthropic/claude-opus-4-8

Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs

Scores

15 sourced entries
HarnessSourceConfidenceLink
Kimi K2.585.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Qwen 3.583.6N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Mistral Large82.8N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Gemini 3.1 Pro81.3N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
MiMo-V2-Flash80.6N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Grok 379.4N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Claude Opus 4.676.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
DeepSeek V3.274.1N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Claude Sonnet 4.672.4N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Gemini 3 Pro71.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
DeepSeek R165.9N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
MiniMax M2.565.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
GPT-oss 120B60.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
GLM-552.0N/ASelf-reportedlowJun 1, 2026livecodebench.github.io
Llama 4 Maverick43.4N/ASelf-reportedlowJun 1, 2026livecodebench.github.io

Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.

Reproduced by the community

All runs
Loading community runs…

About

LiveCodeBench (Live Code Bench) — 持续更新竞赛编程题. Scoring: Pass@1. Difficulty: 高. Imported from benchmark_research.md (2026-06).