Vals AI

LiveCodeBench

Our Implementation of the LiveCodeBench benchmark

In index
Coding
Best in coding
Source ↗

Models141

Top result90.5%Claude Fable 5.1

Top-3 spread1.0 pts

Updated1 Sep 2026

Result and cost

139 of 142 results

Results

142 results

#ModelResultCostTime

01Claude Fable 5.190.5% ±0.9$0.26160 s

02Claude Fable 589.8% ±0.9$0.4282 min

03Gemini 3.8 Flash89.5% ±0.9$0.07124 s

04Claude Opus 589.0% ±0.9$0.13459 s

05Gemini 3.7 Flash88.7% ±0.9$0.0415 s

06Gemini 3.1 Pro Preview88.5% ±0.9$0.10189 s

07Grok 4.688.2% ±0.9$0.042 min

08Gemini 3.6 Flash88.1% ±0.9$0.05632 s

09GPT-5.2-Codex88.0% ±0.9$0.1173 min

10Qwen3.8 Max87.9% ±1.0$0.0916 min

11Claude Opus 4.887.8% ±0.9$0.3273 min

12Gemini 3.5 Flash87.6% ±0.9$0.08664 s

13DeepSeek V4 Pro 081387.5% ±1.0$0.0645 min

13DeepSeek V4 Pro 042387.5% ±1.0$0.1112 min

15Grok 4.587.4% ±1.0$0.04510 min

16GPT-5.3 Codex87.3% ±1.0$0.0922 min

16DeepSeek V4 Flash 073187.3% ±1.0$0.022 min

18Kimi K387.2% ±1.0$0.1043 min

19Qwen3.7 Max87.1% ±1.0$0.0873 min

20Kimi K2.686.8% ±1.0$0.0659 min

21GPT-5 mini86.6% ±1.0$0.01234 s

22GPT-5.186.5% ±1.0$0.1262 min

23Gemini 3 Pro86.4% ±1.0$0.12787 s

24Qwen3.6 Plus86.0% ±1.0$0.0788 min

25Inkling-Small85.9% ±1.0$0.0215 min

25GPT-5.6 Terra85.9% ±1.0$0.03144 s

25GPT-585.9% ±1.0$0.0643 min

25Muse Spark 1.185.9% ±1.0$0.0462 min

29Gemini 3 Flash Preview85.6% ±1.0$0.0512 min

30GPT-5.1-Codex85.5% ±1.0$0.1384 min

30Inkling85.5% ±1.0$0.1534 min

32GPT-5.285.4% ±1.0$0.1353 min

33Qwen3.5 Plus Thinking85.3% ±1.0$0.0484 min

33GPT-5.585.3% ±1.0$0.0673 min

35Claude Opus 4.785.1% ±1.0$0.3518 min

36GPT-5 Codex84.7% ±1.0$0.0972 min

36Claude Opus 4.6 (Adaptive)84.7% ±1.0$0.3435 min

38Grok 4.384.5% ±1.0$0.0254 min

39Grok 4.20 (Reasoning)84.3% ±1.0$0.05276 s

40GPT-5.484.1% ±1.0$0.3293 min

41GPT-5.4 nano84.0% ±1.0$0.002560 s

41Qwen3.8-27B84.0% ±1.0$0.0452 min

41Ling 3.0 Flash84.0% ±1.0$0.00360 s

44o383.9% ±1.0$0.03364 s

44Kimi K2.5 Thinking83.9% ±1.0$0.0393 min

46Claude Opus 4.5 Thinking83.7% ±1.0$0.9713 min

47GPT-5.1-Codex-Max83.6% ±1.1—66 s

48Qwen3.5 Flash83.3% ±1.0$0.00953 min

49Grok 483.2% ±1.0$0.1564 min

49GPT-OSS 120B83.2% ±1.1$0.008582 s

51GPT-5.6 Sol82.6% ±1.1$0.0957 s

52Claude Sonnet 582.4% ±1.1$0.09577 s

53GLM-4.782.2% ±1.0$0.0357 min

53o4-mini82.2% ±1.0$0.01433 s

53MiniMax M382.2% ±1.1$0.0376 min

56Claude Sonnet 4.682.1% ±1.1$0.3715 min

57Kimi K2.7 Code82.0% ±1.1$0.0274 min

58GLM-5 Thinking81.9% ±1.1$0.0445 min

59MiniMax M2.181.8% ±1.1$0.0184 min

60MiMo-V2.581.5% ±1.1$0.00222 min

60GPT-5.4 mini81.5% ±1.1$0.0333 min

62GLM-5.181.4% ±1.1$0.0464 min

62MiMo-V2.5-Pro81.4% ±1.1$0.00685 min

64GLM-4.681.0% ±1.1$0.0274 min

65DeepSeek V3.2 (Thinking)80.7% ±1.1$0.00334 min

66Grok 4.1 Fast (Reasoning)80.6% ±1.1$0.00392 min

67GLM-5.380.5% ±1.1$0.0694 min

67GLM-5.3-Flash80.5% ±1.1$0.00715 min

69GPT-OSS 20B80.4% ±1.1$0.00672 min

70Gemini 3.1 Flash Lite Preview80.1% ±1.1$0.004431 s

71MiniMax M2.779.9% ±1.1$0.0215 min

72MiniMax M2.579.2% ±1.1$0.0335 min

72Gemini 2.5 Pro Preview 03.2579.2% ±1.1$0.023 min

74Gemini 3.5 Flash-Lite79.0% ±1.1$0.01921 s

74Grok 4 Fast (Reasoning)79.0% ±1.1$0.004551 s

76Qwen3 Maxqwen3-max78.2% ±1.1$0.0212 min

77Grok 3 Mini Fast High Reasoning76.2% ±1.1$0.00814 min

77Gemini 2.5 Flash Preview Thinking76.2% ±1.1$0.002449 s

79Gemini 2.5 Flash (Preview)75.1% ±1.1$0.002447 s

80Claude Opus 4.575.0% ±1.1$0.07714 s

81Magistral Medium74.9% ±1.1$0.053 min

82Claude Sonnet 4.5 Thinking73.0% ±1.1$0.1352 min

83Magistral Small72.1% ±1.1$0.0152 min

84o3-mini71.5% ±1.2$0.03354 s

85Gemini 2.5 Flash Lite Preview Thinking71.4% ±1.1$0.0004139 s

86Qwen3 235B-A22B70.6% ±1.1$0.0117 min

87Kimi K2 Instruct70.4% ±1.2$0.006967 s

88DeepSeek-R170.2% ±1.1$0.04186 s

88GPT-5 nano70.2% ±0.9$0.00512 min

88Claude Opus 4 Thinking70.2% ±1.1$0.3972 min

91DeepSeek V3.269.9% ±1.2$0.0006839 s

92GLM-5.269.5% ±1.2$0.1299 min

93Laguna M.168.1% ±1.1$0.00574 min

94Laguna XS.267.8% ±1.1$0.00233 min

95Gemini 2.5 Flash Lite Preview67.7% ±1.2$0.00217 s

96GLM-4.567.4% ±1.1$0.0187 min

97Qwen3 Maxqwen3-max-preview66.9% ±1.2$0.0142 min

98Claude Opus 4.1 Thinking66.5% ±1.1$0.3612 min

99Grok 3 Mini Fast Low Reasoning66.3% ±1.1$0.002231 s

100DeepSeek V3 032465.5% ±1.2$0.001827 s

101Claude Opus 4.164.6% ±1.1$0.05633 s

102Kimi K2 Thinking63.1% ±1.2$0.03419 min

103Claude Opus 462.6% ±1.1$0.05615 s

104Claude Sonnet 4 Thinking62.4% ±1.1$0.04243 s

105Grok Code Fast 162.0% ±1.2$0.008526 s

106Claude 3.7 Sonnet Thinking60.4% ±1.1$0.1252 min

107Claude Sonnet 459.7% ±1.2$0.01414 s

108Llama 3.3 Nemotron Super 49B V1 42e84561 Thinking58.4% ±1.2—3 min

109GPT-4.1 mini58.2% ±1.2$0.001228 s

110Gemini 2.5 Flash Preview 04.1756.9% ±1.2$0.01216 s

111Claude 3.7 Sonnet56.7% ±1.1$0.02116 s

112Mistral Large 355.3% ±1.2$0.002229 s

113GPT-4.154.7% ±1.1$0.009331 s

114Grok 3 [Beta]52.9% ±1.1$0.00945.5 s

115Devstral 2 251251.8% ±1.2$0.05788 s

116o150.3% ±1.4$0.3282 min

117Claude 3.5 Sonnet49.6% ±1.1$0.01211 s

118Llama4 Maverick Instruct Basic47.3% ±1.2$0.000839.4 s

119Gemini 2.5 Flash Preview 04.17 Thinking46.9% ±1.4$0.0022 min

120Grok 4 Fast Non Reasoning46.1% ±1.1$0.0019.0 s

121Mistral Medium44.8% ±1.1$0.001612 s

122Gemini 2.0 Flash 00143.6% ±1.1$0.00033.4 s

123GPT-4o43.4% ±1.1$0.0044.3 s

124Labs Devstral Small43.2% ±1.2$0.0003519 s

125GPT-4.1 nano42.7% ±1.2$0.000312 s

126Grok 4.1 Fast Non Reasoning42.6% ±1.1$0.0004412 s

127Claude 3.5 Haiku41.9% ±1.1$0.003111 s

128Gemini 1.5 Pro 00241.7% ±1.1$0.0024.0 s

129Claude Haiku 4.5 Thinking41.2% ±1.1$0.0682 s

130Grok 238.7% ±1.0$0.0032.4 s

131Llama 4 Scout 17B 16E Instruct38.5% ±1.0$0.0006217 s

132Mistral Large 241137.1% ±1.1$0.00335.3 s

133Gemini 1.5 Flash 00236.9% ±1.1$0.000152.5 s

134Llama 3.3 70B Instruct Turbo36.3% ±1.0$0.000742.9 s

134Llama 3.3 Nemotron Super 49B V1 42e8456136.3% ±1.0—44 s

136Command A35.1% ±1.1$0.006612 s

137Mistral Small 3.131.8% ±1.1$0.000154.0 s

138GPT-4o mini26.4% ±1.1$0.00037.4 s

139Jamba Large 1.622.3% ±1.1$0.00324.8 s

140Command R Plus18.2% ±1.0$0.00335.5 s

141Mistral Small 240215.8% ±1.0$0.000385.0 s

142Jamba Mini 1.69.9% ±0.9$0.00021.1 s