benchgap
Coding

LiveCodeBench Pro leaderboard

As of 2026-11-21, the highest measured score on LiveCodeBench Pro is 90.8% by Sakana Fugu-Ultra. 74 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Claude Fable 5.2100.0%estimated ± 6.0 pp, low confidence
2Claude Opus 5.5100.0%estimated ± 6.0 pp, low confidence
3Claude Sonnet 5.598.0%estimated ± 6.0 pp, low confidence
4Claude Fable 5.197.9%estimated ± 6.0 pp, low confidence
5Claude Mythos 597.4%estimated ± 6.0 pp, low confidence
6Claude Fable 597.2%estimated ± 6.0 pp, low confidence
7Claude Opus 596.8%estimated ± 6.0 pp, low confidence
8Sakana Fugu-Ultra90.8%measured
9Claude Opus 4.890.4%estimated ± 6.0 pp, low confidence
10Ling 3.1 Flash90.1%estimated ± 6.0 pp, low confidence
11Qwen3.8 Max89.4%estimated ± 6.0 pp, low confidence
12Beam 1.0 501B-A23B88.3%estimated ± 6.0 pp, low confidence
13Beam 1.0 501B-A23B FP888.1%estimated ± 6.0 pp, low confidence
14Hy4 preview88.0%estimated ± 6.0 pp, low confidence
15Beam 1.0 501B-A23B NVFP487.9%estimated ± 6.0 pp, low confidence
16Sakana Fugu87.8%measured
17Ornith-1.5-397B87.5%estimated ± 6.0 pp, low confidence
18GPT-5.487.5%measured
19Grok 4.587.3%estimated ± 6.0 pp, low confidence
20GPT-5.6 Sol87.2%estimated ± 6.0 pp, low confidence
21Claude Opus 4.787.0%estimated ± 6.0 pp, low confidence
22GPT-5.6 Terra86.3%estimated ± 6.0 pp, low confidence
23Qwen3.8-Omni-Flash86.2%estimated ± 6.0 pp, low confidence
24Claude Sonnet 586.2%estimated ± 6.0 pp, low confidence
25GPT-5.6 Luna85.8%estimated ± 6.0 pp, low confidence
26Qwen3.8-Flash-Next85.6%estimated ± 6.0 pp, low confidence
27Ornith-1.0-397B85.4%estimated ± 6.0 pp, low confidence
28GLM-5.285.3%estimated ± 6.0 pp, low confidence
29Qwen3.8-27B85.0%estimated ± 6.0 pp, low confidence
30Muse Spark 1.184.9%estimated ± 6.0 pp, low confidence
31Ripple-65B Preview84.8%estimated ± 6.0 pp, low confidence
32Sage-A1.584.7%estimated ± 6.0 pp, low confidence
33dots3-note Preview84.5%estimated ± 6.0 pp, low confidence
34Qwen3.7 Max84.2%estimated ± 6.0 pp, low confidence
35Atria Dawn Preview83.4%estimated ± 6.0 pp, low confidence
36Ornith-1.5-35B-A3B83.4%estimated ± 6.0 pp, low confidence
37Laguna S 2.183.3%estimated ± 6.0 pp, low confidence
38MiniMax M383.0%estimated ± 6.0 pp, low confidence
39Gemini 3.1 Pro82.9%measured
40GPT-5.582.6%estimated ± 6.0 pp, low confidence
41Kimi K2.682.6%estimated ± 6.0 pp, low confidence
42GLM-5.182.5%estimated ± 6.0 pp, low confidence
43Qwen3.7 Plus81.8%estimated ± 6.0 pp, low confidence
44Qwen 3.6 Max (preview)81.6%estimated ± 6.0 pp, low confidence
45MiMo-V2.5-Pro81.5%estimated ± 6.0 pp, low confidence
46Claude Opus 4.581.4%estimated ± 6.0 pp, low confidence
47GPT-5.3 Codex81.2%estimated ± 6.0 pp, low confidence
48Ling 3.0 Flash81.0%estimated ± 6.0 pp, low confidence
49Qwen3.6 Plus81.0%estimated ± 6.0 pp, low confidence
50Step 3.7 Flash80.8%estimated ± 6.0 pp, low confidence
51MiniMax M2.780.7%estimated ± 6.0 pp, low confidence
52MiMo-V2.580.6%estimated ± 6.0 pp, low confidence
53Inkling-Small80.5%estimated ± 6.0 pp, low confidence
54GPT-5.280.2%estimated ± 6.0 pp, low confidence
55DeepSeek V4 Pro80.0%estimated ± 6.0 pp, low confidence
56Muse Spark80.0%measured
57Gemini 3.5 Flash79.8%estimated ± 6.0 pp, low confidence
58GLM-579.8%estimated ± 6.0 pp, low confidence
59Inkling79.1%estimated ± 6.0 pp, low confidence
60Spark-X2.6-4B79.1%estimated ± 6.0 pp, low confidence
61Gemini 3.5 Flash-Lite79.0%estimated ± 6.0 pp, low confidence
62Qwen3.6-27B78.4%estimated ± 6.0 pp, low confidence
63MAI-Thinking-177.8%estimated ± 6.0 pp, low confidence
64DeepSeek V4 Flash 073177.7%estimated ± 6.0 pp, low confidence
65Nemotron 3.5 Super VL 120B A12B77.0%estimated ± 6.0 pp, low confidence
66Nemotron 3.5 Super VL 120B A12B NVFP476.5%estimated ± 6.0 pp, low confidence
67Muse Glimmer 30B76.4%estimated ± 6.0 pp, low confidence
68Qwen3.5 397B76.2%estimated ± 6.0 pp, low confidence
69Kimi K2.576.0%estimated ± 6.0 pp, low confidence
70Ornith-1.0-35B75.7%estimated ± 6.0 pp, low confidence
71Qwen3.6-35B-A3B74.9%estimated ± 6.0 pp, low confidence
72Laguna M.174.6%estimated ± 6.0 pp, low confidence
73Grok 4.2074.2%measured
74Laguna XS 2.173.2%estimated ± 6.0 pp, low confidence
75Ornith-1.5-9B73.1%estimated ± 6.0 pp, low confidence
76Laguna XS.271.9%estimated ± 6.0 pp, low confidence
77Claude Opus 4.670.7%measured
78Ornith-1.0-9B68.6%estimated ± 6.0 pp, low confidence
79LongCat-Flash-Lite-Sparse66.2%estimated ± 6.0 pp, low confidence
80Granite 4.2 30B57.9%estimated ± 6.0 pp, low confidence
81LLaDA2.2-flash54.0%estimated ± 6.0 pp, low confidence
82LFM3-1.7B31.7%measured
83MiniCPM5-1B22.7%measured

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General