Best AI models for agents

218 models · Agentic index. Two or more agentic results. Method

Ranked models

November 2026

#ModelAgenticIndexOutput / 1MSpeed

01Gemini 4 Carbon83.282.7$10.0031/s

02Claude Fable 5.282.285.6$40.0064/s

03GPT-6.1 Terra78.677.1$5.0046/s

04Muse Wonder Full78.277.3——

05Claude Sonnet 5.577.577.2$10.00107/s

06Gemini 4 Argon77.378.5$10.0028/s

07Kimi K3.1 Pro76.775.7$10.0043/s

08GPT-6.1 Sol76.078.0$10.0034/s

09GPT-6 Astra74.678.0$50.0034/s

10Hy474.572.0$1.2552/s

11Claude Opus 5.574.178.6$20.0076/s

12Qwen3.8 Max Preview73.564.2——

13Muse Wonder73.073.1$20.0080/s

14DeepSeek V4.1 Flash 101772.268.5$0.40261/s

15MiMo-V2.6-Pro72.170.2$0.8731/s

16GPT-5.6 Sol71.973.0$22.0064/s

16Grok 4.771.969.1$6.0051/s

18MiniMax M3.1 Flash71.869.3$0.60100/s

18DeepSeek V4 Flash Vision Exp71.860.4$1.3274/s

20Claude Opus 571.673.8$25.0048/s

21Muse Ember71.464.8$0.50158/s

21Claude Fable 571.474.0$50.0056/s

23Claude Fable 5.171.176.2$50.0069/s

24Muse Spark 1.370.470.1$4.2576/s

25Claude Mythos 570.070.8$50.00—

26MiMo-V2.6-Flash69.764.3$0.2828/s

26GLM-5.3-Flash69.763.1$0.078122/s

28Step 5 Preview68.868.4$2.70105/s

29GPT-6 Sol68.672.8$10.0061/s

29Claude Haiku 5.568.666.3$0.50159/s

31Qwen3.8-Flash-Next68.564.3—55/s

32Grok 4.668.467.8$6.0056/s

33GLM-5.368.265.8$1.4086/s

33GPT-5.6 Terra68.268.9$12.0066/s

35Inkling 2 Small67.666.4$1.00—

36GPT-6 Luna66.965.5$0.5066/s

37Ling 3.1 Flash66.667.2$0.13581/s

38Claude Opus 4.866.567.7$25.0069/s

39Qwen3.8 Max (0902)66.463.4$6.0033/s

40Mistral Large 466.360.7$2.10113/s

41Kimi K365.966.9$16.0044/s

42GPT-5.565.869.1$30.0052/s

43Gemini 3.8 Flash65.367.8$4.1368/s

43Claude Sonnet 565.364.2$10.0075/s

45Gemini 3.7 Flash65.168.5$3.7585/s

46Muse Spark Open64.865.1$4.2586/s

47GPT-5.6 Luna64.765.5$1.2051/s

47Gemini 3.5 Flash64.764.2$9.00137/s

49Qwen3.8-Omni-Flash64.562.7$0.4741/s

50Qwen3.8 Max63.765.6—41/s

50DeepSeek V4 Pro 081363.765.5$4.4442/s

52Apodex 1.163.558.9——

52Claude Opus 4.7 (Adaptive)63.563.2$25.00—

52Hy4 preview63.566.4$2.5067/s

52Claude Opus 4.6 (Adaptive)63.560.0——

56dots3-note Preview63.363.4Free—

57Muse Spark 1.262.764.9$4.25123/s

58Beam 1.062.358.8——

59GPT-5.462.164.8$15.0038/s

60Muse Spark 1.161.964.6$4.25152/s

61GLM-5.261.760.6$6.80111/s

62Grok 4.561.663.7$6.0053/s

63Claude Opus 4.760.662.0$25.0047/s

63GPT-5.3 Codex60.662.0$14.0052/s

65Qwen3.7 Max60.560.5$4.4368/s

66Mistral Medium 3.5 128B60.150.3$7.50—

67DeepSeek V4 Flash 073159.961.9$0.3061/s

67Apodex 1.1 Mini59.955.7Free—

69Claude Sonnet 4.659.757.0$15.0044/s

69Gemini 3.6 Flash59.763.7$3.75121/s

69Claude Opus 4.5 Thinking59.756.3——

72Step 3.7 Flash59.655.3$1.15112/s

73Kimi K2.659.558.2$3.4181/s

74Qwen3.5 397B (Reasoning)59.455.3$3.60—

74Sage-A1.559.460.5——

76Beam 1.0 501B-A23B59.360.2—345/s

77Qwen 3.6 Max (preview)59.257.2$6.1675/s

77GLM-5.159.256.0$4.40102/s

79Gemini 3.1 Pro58.462.3$12.00151/s

79Kimi K2.7 Code58.457.7$3.5035/s

81MiMo-V2.5-Pro58.256.3$1.8047/s

82GPT-5.2-Codex58.159.9$14.0044/s

83Claude Opus 4.657.858.0$25.0036/s

84GPT-5.1-Codex-Max57.553.8$10.0055/s

85Qwen3.8-27B57.459.2$1.88113/s

85MiniMax M357.455.7$1.2079/s

87Gemini 3 Flash Preview57.357.5$3.0078/s

87Qwen3.7 Plus57.357.7$1.2851/s

89Claude Sonnet 4.5 Thinking57.153.8——

90GLM-5-Turbo57.051.9$4.0025/s

91Qwen3.6-27B56.752.1$3.2060/s

92MiMo-V2.556.554.8$2.004/s

93K-EXAONE 2.056.448.8——

94Gemini 3 Pro56.257.9$12.00109/s

95GPT-5.1-Codex56.152.9$10.0055/s

95GPT-5.4 mini56.154.5$4.5095/s

97DeepSeek V4 Pro 042355.957.6$1.7889/s

97Grok Build 0.155.952.7$2.00147/s

97Muse Spark55.959.8——

100Inkling-Small55.755.1$1.2080/s

101Beam 1.0 501B-A23B FP855.459.2——

102GPT-5.255.359.0$14.0045/s

103Qwen3.6 Plus55.054.6$1.9544/s

103Beam 1.0 501B-A23B NVFP455.058.5——

105Gemini 3.5 Flash-Lite54.752.0$2.50130/s

105Claude Opus 4.554.752.0$25.0039/s

107Agents-A1-4B54.146.8——

108Ternary Bonsai 2 27B54.052.2$0.5021/s

109Ling 3.0 Flash53.550.9$0.06373/s

110Qwen3.5 397B53.353.8$3.6059/s

110MiMo-V2-Pro53.351.8——

110Muse Glimmer 30B53.353.0$1.2090/s

113GPT-553.155.0$10.0076/s

114Qwen3.6-35B-A3B52.750.6$0.9065/s

115Hy3 Preview52.652.3$0.6069/s

116Gemini 3.1 Pro Preview52.561.5$12.0077/s

117GPT-5.152.455.3$10.0050/s

118MiniMax M2.752.051.2$0.8435/s

119GLM-551.651.6$3.2045/s

120Qwen3.5-122B-A10B51.451.9$2.0870/s

121Inkling51.254.8$4.0585/s

122Grok 4.2050.752.8$2.5072/s

123Grok 4.350.653.3$2.5061/s

123Kimi K2.5 (Reasoning)50.653.0$3.00—

125GPT-5.4 nano50.550.9$1.2591/s

125GLM-5V-Turbo50.548.1$4.0030/s

127Qwen3.5-27B50.351.3$2.6069/s

128Claude Haiku 4.5 Thinking50.243.5——

129Solar Pro 449.952.1$0.6044/s

130GPT-5 (high)49.753.6$10.0076/s

131Ripple-65B Preview49.552.3$0.6063/s

132Kimi K2.549.452.2$3.3236/s

133Claude Sonnet 4.549.347.9$15.0045/s

133Mercury 2.549.346.2$0.15125/s

135Grok 4.20 (Reasoning)48.955.5——

135Nemotron 3 Ultra48.950.8Free—

135Claude Haiku 4.548.943.2$5.0088/s

138Spark-X2.6-4B48.848.7——

138GLM-4.748.850.0$2.2028/s

138GLM-4.648.841.8$2.0030/s

141Gemini 3 Flash48.555.0$3.00198/s

141MiMo-V2-Flash48.543.4—129/s

141MiniMax M2.148.549.1$1.2034/s

144Grok 4.1 Fast (Reasoning)48.447.3——

145Trinity-Large-Preview48.342.2$1.00—

145Kimi K2.5 Thinking48.352.6——

147Kimi K2 Thinking48.048.2$2.5048/s

147Grok 448.050.0—78/s

149Solar Open 247.953.3——

150Nemotron 3.5 Super VL 120B A12B47.851.0——

151DeepSeek V3.2 (Thinking)47.749.4$2.19—

152Qwen3.5-35B-A3B47.349.4$1.0043/s

153Mistral Medium 3.547.142.7$7.5092/s

153Qwen3 Max47.145.0$3.9062/s

155Solar Pro 347.039.2$0.60114/s

156Ling 2.6 Flash46.937.0—104/s

156MiMo-V2-Omni46.947.2——

158LFM3-1.7B46.441.5$0.20218/s

159A.X K246.351.9——

159LFM2.5-8B-A1B46.330.6—335/s

161Nemotron 3.5 Super VL 120B A12B NVFP445.150.8——

162Gemma 4 31B44.947.9Free35/s

163Kimi K2 Instruct44.643.2——

164Mercury 344.452.3$0.35295/s

165Trinity-Large-Thinking44.043.5$0.80138/s

166Claude 4 Sonnet43.939.5$15.0040/s

167DeepSeek V3.243.144.2$0.3819/s

167K-Exaone43.142.3——

169Grok 4 Fast (Reasoning)42.945.3——

169Grok Code Fast 142.941.2$1.50172/s

171Nemotron 3.5 Bolt 74B-A9B42.346.4——

172Granite 4.2 30B42.041.6——

173Command A+41.740.2$1.50161/s

174GPT-5 mini41.448.4$2.0078/s

175Gemini 3.1 Flash Lite Preview41.349.6$1.50151/s

176Grok 4.1 Fast41.133.3$0.50138/s

177Gemini 2.5 Pro40.446.6$10.0091/s

178GPT-OSS 120B40.143.4$0.6034/s

179GPT-OSS 20B39.841.7$0.13310/s

180LongCat-Flash-Lite-Sparse39.639.3——

181MiniCPM5-2B38.237.1——

182DeepSeek V3.138.037.9—98/s

183GPT-4.137.838.0$8.0074/s

184Llama 4 Maverick36.934.4$1.0061/s

185Granite 4.2 8B36.837.6$0.15101/s

186Sarvam 105B36.436.7—191/s

187Grok 4.1 Fast Non Reasoning35.536.2——

188GPT-4.1 mini35.334.5$1.6072/s

189LLaDA2.2-mini34.824.3——

190DeepSeek V3 032434.037.1——

190Gemma 4 26B A4B34.044.5Free—

192Mistral Small 433.637.9$0.60114/s

193Nemotron 3 Nano Omni 30B A3B33.438.6—228/s

194Granite 4.2 3B33.333.2——

195Nemotron 3.5 Lightning 30B A3B33.028.1——

196LFM2.5-2.6B32.029.3$0.50189/s

196LFM2.5-24B-A2B32.039.8$0.70280/s

198North Mini Code30.838.0Free70/s

199Magistral Medium30.633.7——

200Gemma 4 12B30.539.0——

201Nemotron 3.5 Lightning 30B A3B NVFP430.237.0——

202Mistral Large 329.734.3$1.5061/s

203Llama4 Maverick Instruct Basic29.436.0——

204DeepSeek-R127.638.9$2.5018/s

205Llama 3.3 70B27.228.2——

206Nemotron 3 Nano 30B26.832.9—270/s

206Sarvam 30B26.830.3——

208GPT-4o26.728.7$10.0066/s

208Mistral Small 3.126.724.2—28/s

210Claude 3.5 Haiku26.624.1——

211DeepSeek V326.029.9$1.1037/s

212Command A25.323.9——

212Gemma 4 E4B25.328.2——

212Gemma 4 E2B25.321.9——

215Magistral Small24.831.4——

216GPT-4.1 nano24.226.3$0.4056/s

217Llama 4 Scout23.626.8$0.3063/s

218Gemma 3 27B21.925.3—31/s