Inference
Vals AI

LiveCodeBench

Our Implementation of the LiveCodeBench benchmark

ModelsModels with a result.136
Top resultBest result on this test.90.5%Claude Fable 5.1
Top-3 spreadPoints from first to third.1.0 pts
UpdatedDate the source changed the results.1 Sept 2026

Result and cost

020406080100$0.0001$0.0003$0.001$0.003$0.01$0.03$0.1$0.3$1$3RESULTUS DOLLARS PER TASK · LOG SCALE

Results

137 results
#ModelResultThe score from the source.CostUS dollars to run one task.TimeTime to run one task.
01Claude Fable 5.190.5%$0.26160 s
02Claude Fable 589.8%$0.4282 min
03Gemini 3.8 Flash89.5%$0.07124 s
04Claude Opus 589.0%$0.13459 s
05Gemini 3.7 Flash88.7%$0.0415 s
06Gemini 3.1 Pro Preview88.5%$0.10189 s
07Grok 4.688.2%$0.042 min
08Gemini 3.6 Flash88.1%$0.05632 s
09GPT-5.2-Codex88.0%$0.1173 min
10Qwen3.8 Max87.9%$0.0916 min
11Claude Opus 4.887.8%$0.3273 min
12Gemini 3.5 Flash87.6%$0.08664 s
13DeepSeek V4 Pro 042387.5%$0.0645 min
14Grok 4.587.4%$0.04510 min
15GPT-5.3 Codex87.3%$0.0922 min
16DeepSeek V4 Flash 042387.3%$0.022 min
17Qwen3.7 Max87.1%$0.0873 min
18GPT-5 mini86.6%$0.01234 s
19GPT-5.186.5%$0.1262 min
20Gemini 3 Pro86.4%$0.12787 s
21Qwen3.6 Plus86.0%$0.0788 min
22Inkling-Small85.9%$0.0215 min
23GPT-5.6 Terra85.9%$0.03144 s
24GPT-585.9%$0.0643 min
25Muse Spark 1.185.9%$0.0462 min
26Gemini 3 Flash Preview85.6%$0.0512 min
27GPT-5.1-Codex85.5%$0.1384 min
28Inkling85.5%$0.1534 min
29GPT-5.285.4%$0.1353 min
30Qwen3.5 Plus Thinking85.3%$0.0484 min
31GPT-5.585.3%$0.0673 min
32Claude Opus 4.785.1%$0.3518 min
33GPT-5 Codex84.7%$0.0972 min
34Claude Opus 4.6 (Adaptive)84.7%$0.3435 min
35Grok 4.384.5%$0.0254 min
36Grok 4.20 (Reasoning)84.3%$0.05276 s
37GPT-5.484.1%$0.3293 min
38GPT-5.4 nano84.0%<$0.0160 s
39Qwen3.8-27B84.0%$0.0452 min
40Ling 3.0 Flash84.0%<$0.0160 s
41o383.9%$0.03364 s
42Kimi K2.5 Thinking83.9%$0.0393 min
43Claude Opus 4.5 Thinking83.7%$0.9713 min
44GPT-5.1-Codex-Max83.6%<$0.0166 s
45Qwen3.5 Flash83.3%<$0.013 min
46Grok 483.2%$0.1564 min
47GPT-OSS 120B83.2%<$0.0182 s
48GPT-5.6 Sol82.6%$0.0957 s
49Claude Sonnet 582.4%$0.09577 s
50GLM-4.782.2%$0.0357 min
51o4-mini (high)82.2%$0.01433 s
52MiniMax M382.2%$0.0376 min
53Claude Sonnet 4.682.1%$0.3715 min
54Kimi K2.7 Code82.0%$0.0274 min
55GLM-5 Thinking81.9%$0.0445 min
56MiniMax M2.181.8%$0.0184 min
57MiMo-V2.581.5%<$0.012 min
58GPT-5.4 mini81.5%$0.0333 min
59GLM-5.181.4%$0.0464 min
60MiMo-V2.5-Pro81.4%<$0.015 min
61GLM-4.681.0%$0.0274 min
62DeepSeek V3.2 (Thinking)80.7%<$0.014 min
63Grok 4.1 Fast (Reasoning)80.6%<$0.012 min
64GLM-5.380.5%$0.0694 min
65GLM-5.3-Flash80.5%<$0.015 min
66GPT-OSS 20B80.4%<$0.012 min
67Gemini 3.1 Flash Lite Preview80.1%<$0.0131 s
68MiniMax M2.779.9%$0.0215 min
69MiniMax M2.579.2%$0.0335 min
70Gemini 2.5 Pro Preview 03.2579.2%$0.023 min
71Gemini 3.5 Flash-Lite79.0%$0.01921 s
72Grok 4 Fast (Reasoning)79.0%<$0.0151 s
73Qwen3 Max78.2%$0.0212 min
74Grok 3 Mini Fast High Reasoning76.2%<$0.014 min
75Gemini 2.5 Flash Preview Thinking76.2%<$0.0149 s
76Gemini 2.5 Flash (Preview)75.1%<$0.0147 s
77Claude Opus 4.575.0%$0.07714 s
78Magistral Medium74.9%$0.053 min
79Claude Sonnet 4.5 Thinking73.0%$0.1352 min
80Magistral Small72.1%$0.0152 min
81o3-mini71.5%$0.03354 s
82Gemini 2.5 Flash Lite Preview Thinking71.4%<$0.0139 s
83Qwen3 235B A22B70.6%$0.0117 min
84Kimi K2 Instruct70.4%<$0.0167 s
85DeepSeek-R170.2%$0.04186 s
86GPT-5 nano70.2%<$0.012 min
87Claude Opus 4 Thinking70.2%$0.3972 min
88DeepSeek V3.269.9%<$0.0139 s
89GLM-5.269.5%$0.1299 min
90Laguna M.168.1%<$0.014 min
91Laguna XS.267.8%<$0.013 min
92Gemini 2.5 Flash Lite Preview67.7%<$0.0117 s
93GLM-4.567.4%$0.0187 min
94Qwen3 Max66.9%$0.0142 min
95Claude Opus 4.1 Thinking66.5%$0.3612 min
96Grok 3 Mini Fast Low Reasoning66.3%<$0.0131 s
97DeepSeek V365.5%<$0.0127 s
98Claude Opus 4.164.6%$0.05633 s
99Kimi K2 Thinking63.1%$0.03419 min
100Claude Opus 462.6%$0.05615 s
101Claude Sonnet 4 Thinking62.4%$0.04243 s
102Grok Code Fast 162.0%<$0.0126 s
103Claude 3.7 Sonnet Thinking60.4%$0.1252 min
104Claude Sonnet 459.7%$0.01414 s
105Llama 3.3 Nemotron Super 49B V1 42e84561 Thinking58.4%—3 min
106GPT-4.1 mini58.2%<$0.0128 s
107Gemini 2.5 Flash Preview 04.1756.9%$0.01216 s
108Claude 3.7 Sonnet56.7%$0.02116 s
109GPT-4.154.7%<$0.0131 s
110Grok 3 [Beta]52.9%<$0.016 s
111Devstral51.8%$0.05788 s
112o150.3%$0.3282 min
113Claude 3.5 Sonnet49.6%$0.01211 s
114Llama4 Maverick Instruct Basic47.3%<$0.019 s
115Gemini 2.5 Flash Preview 04.17 Thinking46.9%<$0.012 min
116Mistral Large46.2%<$0.0129 s
117Grok 4 Fast Non Reasoning46.1%<$0.019 s
118Mistral Medium44.8%<$0.0112 s
119Gemini 2.0 Flash 00143.6%<$0.013 s
120GPT-4o43.4%<$0.014 s
121Labs Devstral Small43.2%<$0.0119 s
122GPT-4.1 nano42.7%<$0.0112 s
123Grok 4.1 Fast Non Reasoning42.6%<$0.0112 s
124Claude 3.5 Haiku41.9%<$0.0111 s
125Gemini 1.5 Pro 00241.7%<$0.014 s
126Claude Haiku 4.5 Thinking41.2%$0.0682 s
127Grok 238.7%<$0.012 s
128Llama 4 Scout 17B 16E Instruct38.5%<$0.0117 s
129Gemini 1.5 Flash 00236.9%<$0.013 s
130Llama 3.3 70B Instruct Turbo36.3%<$0.013 s
131Llama 3.3 Nemotron Super 49B V1 42e8456136.3%—44 s
132Command A35.1%<$0.0112 s
133GPT-4o mini26.4%<$0.017 s
134Mistral Small23.8%<$0.014 s
135Jamba Large 1.622.3%<$0.015 s
136Command R Plus18.2%<$0.015 s
137Jamba Mini 1.69.9%<$0.011 s