← LiveBench leaderboards

Coding

LiveBench · 28 models

RankModelScore
1gpt-5.2-codex83.6195
2claude-fable-5-xhigh-effort82.5325
3gpt-5.5-xhigh82.1495
4claude-opus-4-7-xhigh-effort82.088
5claude-sonnet-5-xhigh-effort80.68
6gpt-5.5-high80.037
7claude-opus-4-5-20251101-thinking-64k-high-effort79.654
7glm-5.279.654
9claude-opus-4-8-xhigh-effort79.2715
9claude-sonnet-4-6-thinking-auto-medium-effort79.2715
11gpt-5.578.6285
12kimi-k2.6-thinking78.567
13claude-opus-4-6-thinking-auto-high-effort78.1845
13gemini-3.5-flash-high78.1845
13qwen3.6-plus78.1845
16gpt-5.4-xhigh77.5415
17gemini-3.1-pro-preview-high76.4545
18gpt-5.2-2025-12-11-high76.0715
19qwen3.7-max74.219
20kimi-k2.7-code73.959
21qwen3.6-27b71.785
22gpt-5.4-mini-xhigh71.624
23gpt-5.4-nano-xhigh70.8385
24deepseek-v4-pro69.994
25grok-4.369.9325
26deepseek-v4-flash69.228
27minimax-m368.2025
28grok-build-0.165.3855