最佳编程 AI 模型

209 个模型 · 编程指数。至少两项编程结果。方法

上榜模型

2026年11月

#模型编程指数输出 / 1M速度

01Claude Fable 5.286.285.6$40.0064/s

02Claude Opus 5.581.878.6$20.0076/s

03Claude Sonnet 5.580.877.2$10.00107/s

04Gemini 4 Carbon80.582.7$10.0031/s

05Claude Fable 5.177.776.2$50.0069/s

06Muse Wonder Full77.377.3——

07GPT-6 Astra76.978.0$50.0034/s

08Gemini 4 Argon76.778.5$10.0028/s

09GPT-6.1 Sol75.878.0$10.0034/s

10Claude Opus 574.873.8$25.0048/s

11Claude Fable 574.774.0$50.0056/s

12Claude Mythos 574.470.8$50.00—

12GPT-6.1 Terra74.477.1$5.0046/s

14GPT-6 Sol73.972.8$10.0061/s

15Muse Wonder72.373.1$20.0080/s

16Kimi K3.1 Pro71.975.7$10.0043/s

17Step 5 Preview71.468.4$2.70105/s

18Claude Haiku 5.571.066.3$0.50159/s

19GPT-5.6 Sol70.473.0$22.0064/s

20Grok 4.769.969.1$6.0051/s

21DeepSeek V4.1 Flash 101769.868.5$0.40261/s

22MiniMax M3.1 Flash69.469.3$0.60100/s

23Muse Spark 1.368.870.1$4.2576/s

24GPT-5.6 Terra68.468.9$12.0066/s

24Hy468.472.0$1.2552/s

26Claude Opus 4.868.367.7$25.0069/s

26Sakana Fugu-Ultra68.366.3——

28Grok 4.667.767.8$6.0056/s

29GPT-5.6 Luna67.465.5$1.2051/s

30Gemini 3.7 Flash66.968.5$3.7585/s

30Ling 3.1 Flash66.967.2$0.13581/s

32GPT-6 Luna66.865.5$0.5066/s

33MiMo-V2.6-Pro66.670.2$0.8731/s

33Inkling 2 Small66.666.4$1.00—

35Kimi K366.266.9$16.0044/s

35Gemini 3.8 Flash66.267.8$4.1368/s

37Grok 4.565.863.7$6.0053/s

38GLM-5.365.765.8$1.4086/s

39GPT-5.565.569.1$30.0052/s

39Hy4 preview65.566.4$2.5067/s

41Claude Sonnet 565.364.2$10.0075/s

42DeepSeek V4 Pro 081365.165.5$4.4442/s

43Muse Spark 1.164.864.6$4.25152/s

44Claude Opus 4.7 (Adaptive)64.763.2$25.00—

44Sakana Fugu64.763.9——

46Claude Opus 4.764.562.0$25.0047/s

47Muse Spark 1.264.364.9$4.25123/s

47Qwen3.8 Max64.365.6—41/s

49MiMo-V2.6-Flash64.164.3$0.2828/s

50GLM-5.263.560.6$6.80111/s

51Qwen3.8 Max Preview63.464.2——

52Beam 1.063.358.8——

52Muse Spark Open63.365.1$4.2586/s

54DeepSeek V4 Flash 073162.861.9$0.3061/s

55Mistral Large 462.760.7$2.10113/s

56GLM-5.3-Flash62.363.1$0.078122/s

57Qwen3.8-Flash-Next62.164.3—55/s

58Beam 1.0 501B-A23B62.060.2—345/s

58Gemini 3.6 Flash62.063.7$3.75121/s

60GPT-5.461.864.8$15.0038/s

60GPT-5.3 Codex61.862.0$14.0052/s

62Qwen3.8-Omni-Flash61.562.7$0.4741/s

63Beam 1.0 501B-A23B FP861.459.2——

63GPT-5.2-Codex61.459.9$14.0044/s

65GPT-5.261.059.0$14.0045/s

66Beam 1.0 501B-A23B NVFP460.858.5——

66Gemini 3.5 Flash60.864.2$9.00137/s

68Kimi K2.7 Code60.457.7$3.5035/s

69Qwen3.8-27B60.359.2$1.88113/s

69dots3-note Preview60.363.4Free—

71Claude Opus 4.6 (Adaptive)60.060.0——

72Qwen3.7 Max59.860.5$4.4368/s

73Kimi K2.659.358.2$3.4181/s

74Claude Opus 4.659.258.0$25.0036/s

74DeepSeek V4 Pro 042359.257.6$1.7889/s

76Claude Sonnet 4.658.957.0$15.0044/s

77Gemini 3.1 Pro Preview58.861.5$12.0077/s

78Qwen3.7 Plus58.357.7$1.2851/s

79Gemini 3.1 Pro58.062.3$12.00151/s

79Solar Pro 458.052.1$0.6044/s

81MiniMax M357.755.7$1.2079/s

82Sage-A1.557.360.5——

83MiMo-V2.5-Pro57.256.3$1.8047/s

84Qwen 3.6 Max (preview)57.057.2$6.1675/s

84GLM-5.157.056.0$4.40102/s

86Apodex 1.156.858.9——

87Solar Open 256.753.3——

88Gemini 3 Pro56.457.9$12.00109/s

89Qwen3.6 Plus56.254.6$1.9544/s

90Kimi K2.5 (Reasoning)56.053.0$3.00—

91MiMo-V2.555.554.8$2.004/s

92Grok 4.2055.452.8$2.5072/s

93Ripple-65B Preview55.352.3$0.6063/s

94Hy3 Preview55.252.3$0.6069/s

94MAI-Thinking-155.253.5——

96GPT-5.4 mini55.154.5$4.5095/s

96Nemotron 3.5 Super VL 120B A12B NVFP455.150.8——

98Claude Opus 4.555.052.0$25.0039/s

98Gemini 3 Flash Preview55.057.5$3.0078/s

100Inkling-Small54.955.1$1.2080/s

100Step 3.7 Flash54.955.3$1.15112/s

102Claude Opus 4.5 Thinking54.456.3——

102Muse Glimmer 30B54.453.0$1.2090/s

102Gemini 3.5 Flash-Lite54.452.0$2.50130/s

105Nemotron 3 Ultra54.350.8Free—

106Inkling54.154.8$4.0585/s

107Qwen3.5 Plus Thinking53.948.1——

107Kimi K2.553.952.2$3.3236/s

109GPT-5.4 nano53.750.9$1.2591/s

110GLM-553.551.6$3.2045/s

110GPT-5.153.555.3$10.0050/s

112Mistral Medium 3.5 128B53.450.3$7.50—

113Muse Spark53.359.8——

113Qwen3.5 397B53.353.8$3.6059/s

115GPT-5.1-Codex-Max53.253.8$10.0055/s

115Grok 4.353.253.3$2.5061/s

115GLM-5 Thinking53.253.4——

115GLM-4.753.250.0$2.2028/s

119Nemotron 3.5 Super VL 120B A12B53.051.0——

119MiMo-V2-Flash53.043.4—129/s

121Qwen3.6-27B52.852.1$3.2060/s

122Qwen3.5-27B52.751.3$2.6069/s

123GPT-552.555.0$10.0076/s

124MiniMax M2.752.351.2$0.8435/s

125Ternary Bonsai 2 27B52.252.2$0.5021/s

126Claude 4 Sonnet52.139.5$15.0040/s

127Kimi K2.5 Thinking51.652.6——

127MiniMax M2.151.649.1$1.2034/s

127Grok 4.20 (Reasoning)51.655.5——

130A.X K251.351.9——

131o351.051.8$8.0065/s

132Grok 450.950.0—78/s

133MiniMax M2.550.750.2$1.0844/s

134Claude Sonnet 4.5 Thinking50.653.8——

135Qwen3.5-122B-A10B50.451.9$2.0870/s

136Gemma 4 31B50.147.9Free35/s

137Ling 3.0 Flash50.050.9$0.06373/s

138Claude Haiku 4.549.943.2$5.0088/s

139GPT-5.1-Codex49.852.9$10.0055/s

140Qwen3.6-35B-A3B49.650.6$0.9065/s

140K-EXAONE 2.049.648.8——

142GPT-5 mini49.448.4$2.0078/s

143DeepSeek V3.2 (Thinking)49.249.4$2.19—

144Qwen3.5-35B-A3B49.049.4$1.0043/s

145GPT-OSS 20B48.941.7$0.13310/s

146Gemini 2.5 Pro48.846.6$10.0091/s

146Spark-X2.6-4B48.848.7——

146DeepSeek V3.248.844.2$0.3819/s

149Gemini 3.1 Flash Lite Preview48.349.6$1.50151/s

149Trinity-Large-Thinking48.343.5$0.80138/s

151Claude Opus 4.148.247.5$75.0012/s

152LongCat-Flash-Lite-Sparse47.539.3——

153Claude Opus 447.343.9——

154Mistral Medium 3.547.142.7$7.5092/s

155DeepSeek-R1-052846.944.6$2.1520/s

155Grok Build 0.146.952.7$2.00147/s

157GLM-4.646.741.8$2.0030/s

158Qwen3 Max46.345.0$3.9062/s

159Claude Sonnet 446.243.0$15.0049/s

160Grok Code Fast 145.541.2$1.50172/s

161Gemma 4 26B A4B45.344.5Free—

162o4-mini45.244.7$4.40122/s

163Kimi K2 Thinking44.948.2$2.5048/s

163Grok 4 Fast (Reasoning)44.945.3——

165Mercury 2.544.546.2$0.15125/s

166Mistral Small 444.337.9$0.60114/s

167Qwen3 235B-A22B44.038.1$1.8244/s

168GLM-4.543.345.7$2.2025/s

168Grok 4.1 Fast (Reasoning)43.347.3——

170Granite 4.2 30B43.241.6——

171DeepSeek V4 Flash Vision Exp43.160.4$1.3274/s

172DeepSeek V3.143.037.9—98/s

173GPT-OSS 120B42.843.4$0.6034/s

174DeepSeek-R142.638.9$2.5018/s

175Claude Haiku 4.5 Thinking42.543.5——

176Claude 3.7 Sonnet42.137.7——

177Claude Sonnet 4 Thinking41.846.9——

178Kimi K2 Instruct41.343.2——

179DeepSeek V3 032440.937.1——

180Magistral Medium39.233.7——

181Kimi K239.039.6$2.3037/s

182Nemotron 3.5 Lightning 30B A3B NVFP438.737.0——

182LFM2.5-24B-A2B38.739.8$0.70280/s

184Mistral Large 338.234.3$1.5061/s

185GPT-5 nano37.241.9$0.4081/s

186o3-mini37.140.0$4.40173/s

187LFM3-1.7B37.041.5$0.20218/s

188Granite 4.2 8B36.537.6$0.15101/s

189Magistral Small36.331.4——

190Granite 4.2 3B34.933.2——

191Grok 4.1 Fast Non Reasoning34.336.2——

192GPT-4.134.238.0$8.0074/s

193Claude 3.5 Sonnet34.032.2$15.00—

194Agents-A1-4B33.946.8——

195MiniCPM5-2B32.037.1——

196Gemini 2.5 Flash31.935.0$2.5095/s

197DeepSeek V331.129.9$1.1037/s

198Sarvam 30B29.430.3——

199GPT-4.1 mini28.234.5$1.6072/s

200Claude 3.5 Haiku28.024.1——

201Gemma 4 E4B27.428.2——

202Mistral Small 3.125.624.2—28/s

203Grok 225.229.1——

204LFM2.5-2.6B23.929.3$0.50189/s

205GPT-4o22.828.7$10.0066/s

206Nemotron 3.5 Lightning 30B A3B21.928.1——

207GPT-4o mini19.526.2$0.6056/s

208Command A12.723.9——

209Claude 3 Opus9.523.3$75.00—