benchgap
Coding

LiveCodeBench leaderboard

As of 2026-11-21, the highest measured score on LiveCodeBench is 87.8% by Solar Pro 4. 107 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Claude Opus 588.1%estimated ± 14.6 pp, low confidence
2Claude Mythos 587.9%estimated ± 14.6 pp, low confidence
3Solar Pro 487.8%measured
4Claude Fable 587.6%estimated ± 14.6 pp, low confidence
5Ember-186.2%estimated ± 14.6 pp, low confidence
6Claude Opus 4.884.2%estimated ± 14.6 pp, low confidence
7Claude Opus 4.783.7%estimated ± 14.6 pp, low confidence
8Ornith-1.5-397B82.8%estimated ± 14.6 pp, low confidence
9Claude Sonnet 582.3%estimated ± 14.6 pp, low confidence
10GPT-5.3 Codex82.2%estimated ± 14.6 pp, low confidence
11Ornith-1.0-397B80.7%estimated ± 14.6 pp, low confidence
12Inkling 2 Small80.6%estimated ± 14.6 pp, low confidence
13Beam 1.0 501B-A23B80.2%estimated ± 14.6 pp, low confidence
14Claude Opus 4.579.8%estimated ± 14.6 pp, low confidence
15Claude Opus 4.679.8%estimated ± 14.6 pp, low confidence
16Beam 1.0 501B-A23B NVFP479.7%estimated ± 14.6 pp, low confidence
17DeepSeek V4 Pro79.6%estimated ± 14.6 pp, low confidence
18MiniMax M379.6%estimated ± 14.6 pp, low confidence
19Qwen3.7 Max79.5%estimated ± 14.6 pp, low confidence
20Inkling-Small79.4%estimated ± 14.6 pp, low confidence
21Kimi K2.679.4%estimated ± 14.6 pp, low confidence
22Beam 1.0 501B-A23B FP879.4%estimated ± 14.6 pp, low confidence
23GPT-5.279.3%estimated ± 14.6 pp, low confidence
24Claude Sonnet 4.679.0%estimated ± 14.6 pp, low confidence
25Inkling 2 Mini78.8%estimated ± 14.6 pp, low confidence
26DeepSeek V4 Flash 073178.7%estimated ± 14.6 pp, low confidence
27Ornith-1.5-35B-A3B78.7%estimated ± 14.6 pp, low confidence
28Qwen3.6 Plus78.6%estimated ± 14.6 pp, low confidence
29BTL-478.3%estimated ± 14.6 pp, low confidence
30dots3-note Preview78.3%estimated ± 14.6 pp, low confidence
31MiMo-V2-Pro78.1%estimated ± 14.6 pp, low confidence
32GLM-577.9%estimated ± 14.6 pp, low confidence
33Apodex 1.177.9%estimated ± 14.6 pp, low confidence
34Qwen3.7 Plus77.9%estimated ± 14.6 pp, low confidence
35Inkling77.8%estimated ± 14.6 pp, low confidence
36Mistral Medium 3.5 128B77.8%estimated ± 14.6 pp, low confidence
37Muse Spark77.7%estimated ± 14.6 pp, low confidence
38Claude Sonnet 4.577.6%estimated ± 14.6 pp, low confidence
39Qwen3.6-27B77.6%estimated ± 14.6 pp, low confidence
40Kimi K2.577.3%estimated ± 14.6 pp, low confidence
41Grok 4.2077.3%estimated ± 14.6 pp, low confidence
42Ripple-65B Preview77.1%estimated ± 14.6 pp, low confidence
43Qwen3.5 397B77.0%estimated ± 14.6 pp, low confidence
44Muse Glimmer 30B76.8%estimated ± 14.6 pp, low confidence
45Ornith-1.0-35B76.6%estimated ± 14.6 pp, low confidence
46MiMo-V2-Omni76.1%estimated ± 14.6 pp, low confidence
47Laguna M.176.0%estimated ± 14.6 pp, low confidence
48Claude 4.1 Opus75.9%estimated ± 14.6 pp, low confidence
49Hy3 Preview75.8%estimated ± 14.6 pp, low confidence
50Step 3.5 Flash75.8%estimated ± 14.6 pp, low confidence
51Flint75.7%estimated ± 14.6 pp, low confidence
52GLM-4.775.4%estimated ± 14.6 pp, low confidence
53MAI-Thinking-175.3%estimated ± 14.6 pp, low confidence
54MiMo-V2-Flash75.2%estimated ± 14.6 pp, low confidence
55Qwen3.6-35B-A3B75.2%estimated ± 14.6 pp, low confidence
56Claude Haiku 4.575.1%estimated ± 14.6 pp, low confidence
57Jais-3-MoVA-36B-A4B74.8%estimated ± 14.6 pp, low confidence
58Grok 3 Mini74.8%measured
59Claude 4 Sonnet74.7%estimated ± 14.6 pp, low confidence
60MAI-Code-1.1-Flash74.7%estimated ± 14.6 pp, low confidence
61Claude 4 Opus74.6%estimated ± 14.6 pp, low confidence
62Jais-3-32B74.6%estimated ± 14.6 pp, low confidence
63Qwen3.5-27B74.6%estimated ± 14.6 pp, low confidence
64Qwen3.5-122B-A10B74.3%estimated ± 14.6 pp, low confidence
65Nemotron 3 Ultra74.2%estimated ± 14.6 pp, low confidence
66Nemotron 3.5 Super VL 120B A12B73.7%estimated ± 14.6 pp, low confidence
67Nemotron 3.5 Super VL 120B A12B NVFP473.7%estimated ± 14.6 pp, low confidence
68Laguna XS 2.173.6%estimated ± 14.6 pp, low confidence
69Grok Code Fast 173.5%estimated ± 14.6 pp, low confidence
70Ornith-1.5-9B73.4%estimated ± 14.6 pp, low confidence
71DeepSeek-R1-052873.3%measured
72Solar Open 273.2%estimated ± 14.6 pp, low confidence
73Jais-3-375B-A23B73.1%estimated ± 14.6 pp, low confidence
74Jais-3-7B73.1%estimated ± 14.6 pp, low confidence
75Laguna XS.272.9%estimated ± 14.6 pp, low confidence
76Qwen3-Coder 480B-A35B72.7%estimated ± 14.6 pp, low confidence
77Ornith-1.0-9B72.6%estimated ± 14.6 pp, low confidence
78Qwen3.5-35B-A3B72.4%estimated ± 14.6 pp, low confidence
79K-EXAONE 2.071.8%estimated ± 14.6 pp, low confidence
80LongCat-Flash-Lite-Sparse71.8%estimated ± 14.6 pp, low confidence
81Kimi K270.1%estimated ± 14.6 pp, low confidence
82Gemini 2.5 Pro68.7%estimated ± 14.6 pp, low confidence
83Spark-X2.6-4B68.0%estimated ± 14.6 pp, low confidence
84GPT-OSS 120B67.7%estimated ± 14.6 pp, low confidence
85Claude 3.7 Sonnet67.7%estimated ± 14.6 pp, low confidence
86Ternary Bonsai 2 27B66.6%estimated ± 14.6 pp, low confidence
87GPT-OSS 20B66.5%estimated ± 14.6 pp, low confidence
88Nemotron 3 Super 100B66.3%estimated ± 14.6 pp, low confidence
89Nemotron Ultra 253B66.3%measured
90GLM-4.7-Flash65.4%estimated ± 14.6 pp, low confidence
91MiniMax M1 80k65.0%measured
92Granite 4.2 30B63.7%estimated ± 14.6 pp, low confidence
93GPT-4.161.9%estimated ± 14.6 pp, low confidence
94ZAYA1-74B-Preview60.8%estimated ± 14.6 pp, low confidence
95DeepSeek-R1-0528-Qwen3-8B60.5%measured
96Nemotron 3.5 Lightning 30B A3B NVFP460.5%estimated ± 14.6 pp, low confidence
97LFM2.5-24B-A2B59.0%estimated ± 14.6 pp, low confidence
98K-Exaone57.7%estimated ± 14.6 pp, low confidence
99o3-mini57.6%estimated ± 14.6 pp, low confidence
100LLaDA2.2-flash57.6%estimated ± 14.6 pp, low confidence
101DeepSeek-R157.5%estimated ± 14.6 pp, low confidence
102DeepSeek R1 Distill Llama 70B57.5%measured
103Claude 3.5 Sonnet57.4%estimated ± 14.6 pp, low confidence
104Grok 3 [Beta]57.0%measured
105DeepSeek V3.156.4%measured
106Granite 4.2 8B56.2%estimated ± 14.6 pp, low confidence
107Mellum2.1-12B-A2.5B-Thinking55.7%estimated ± 14.6 pp, low confidence
108Devstral Small55.5%estimated ± 14.6 pp, low confidence
109MiniCPM5-2B55.2%estimated ± 14.6 pp, low confidence
110Sarvam 105B53.9%estimated ± 14.6 pp, low confidence
111Phi-4-reasoning53.8%measured
112DeepSeek R1 Distill Qwen 14B53.1%measured
113Phi-4-reasoning-plus53.1%measured
114LFM3-1.7B52.5%estimated ± 14.6 pp, low confidence
115Claude 3.5 Haiku50.0%estimated ± 14.6 pp, low confidence
116Sarvam 30B43.6%estimated ± 14.6 pp, low confidence
117Gemini 2.5 Flash-Lite41.2%estimated ± 14.6 pp, low confidence
118Qwen2.5-Max38.7%measured
119DeepSeek V337.6%measured
120GPT-4.1 mini32.5%estimated ± 14.6 pp, low confidence
121Gemini 2.0 Flash29.9%estimated ± 14.6 pp, low confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General