最佳知识类 AI 模型

268 个模型 · 知识指数。至少两项结果。方法

上榜模型

2026年11月

#模型知识指数输出 / 1M速度

01Claude Fable 5.287.185.6$40.0064/s

02Gemini 4 Carbon79.582.7$10.0031/s

03Claude Opus 5.579.278.6$20.0076/s

04GPT-6 Astra78.578.0$50.0034/s

05GPT-6.1 Sol78.078.0$10.0034/s

06Claude Fable 5.177.276.2$50.0069/s

07Gemini 4 Argon76.578.5$10.0028/s

08Claude Fable 575.874.0$50.0056/s

09Claude Mythos 574.970.8$50.00—

10Claude Opus 573.173.8$25.0048/s

11GPT-5.6 Sol72.773.0$22.0064/s

12GPT-6 Sol71.972.8$10.0061/s

13Muse Wonder Full71.777.3——

14Claude Sonnet 5.571.377.2$10.00107/s

15Gemini 3.7 Flash71.268.5$3.7585/s

16Kimi K3.1 Pro71.075.7$10.0043/s

17Gemini 3.8 Flash70.967.8$4.1368/s

18GPT-6.1 Terra70.577.1$5.0046/s

19GPT-5.570.469.1$30.0052/s

19Ling 3.1 Flash70.467.2$0.13581/s

21Sakana Fugu-Ultra69.766.3——

22Gemini 3.1 Pro69.462.3$12.00151/s

23Claude Opus 4.7 (Adaptive)69.163.2$25.00—

24GPT-5.5 Pro69.074.8$180.0019/s

25MiniMax M3.1 Flash68.969.3$0.60100/s

26Muse Wonder68.673.1$20.0080/s

27Gemini 3.5 Flash68.564.2$9.00137/s

28Sakana Fugu68.363.9——

29Hy4 preview67.966.4$2.5067/s

29Gemini 3.6 Flash67.963.7$3.75121/s

31Muse Spark67.859.8——

32Muse Spark 1.367.570.1$4.2576/s

33Muse Spark 1.167.464.6$4.25152/s

34GPT-5.4 Pro67.273.3$180.0021/s

35Grok 4.766.569.1$6.0051/s

36Claude Opus 4.866.467.7$25.0069/s

36GPT-5.466.464.8$15.0038/s

38Kimi K366.366.9$16.0044/s

39Gemini 3 Pro66.157.9$12.00109/s

40Grok 4.565.863.7$6.0053/s

41Grok 4.665.767.8$6.0056/s

42GPT-5.6 Terra65.568.9$12.0066/s

43Muse Spark Open65.465.1$4.2586/s

44Muse Spark 1.265.264.9$4.25123/s

45DeepSeek V4 Pro 081365.065.5$4.4442/s

45Gemini 3.1 Pro Preview65.061.5$12.0077/s

47MiMo-V2.6-Pro64.270.2$0.8731/s

47DeepSeek V4.1 Flash 101764.268.5$0.40261/s

49GPT-5.3 Codex64.162.0$14.0052/s

50Step 5 Preview63.568.4$2.70105/s

50GPT-6 Luna63.565.5$0.5066/s

52Qwen3.8 Max62.665.6—41/s

52Claude Opus 4.6 (Adaptive)62.660.0——

52Qwen3.8 Max (0902)62.663.4$6.0033/s

55GPT-5.6 Luna62.165.5$1.2051/s

56Claude Opus 4.661.858.0$25.0036/s

56Claude Opus 4.761.862.0$25.0047/s

58Inkling 2 Small61.566.4$1.00—

59MiniMax M361.055.7$1.2079/s

60Qwen3.8-Omni-Flash60.862.7$0.4741/s

61Gemini 3 Flash60.755.0$3.00198/s

62GPT-5.260.659.0$14.0045/s

63Beam 1.0 501B-A23B FP860.359.2——

64GPT-5.2-Codex60.159.9$14.0044/s

65DeepSeek V4 Pro 042359.957.6$1.7889/s

66Qwen 3.6 Max (preview)59.357.2$6.1675/s

66Qwen3.8 Max Preview59.364.2——

66GLM-5.359.365.8$1.4086/s

69Beam 1.0 501B-A23B NVFP459.258.5——

70Claude Opus 4.5 Thinking59.056.3——

70Claude Sonnet 559.064.2$10.0075/s

72Kimi K2.7 Code58.857.7$3.5035/s

73DeepSeek V4 Flash 073158.661.9$0.3061/s

73Qwen3.7 Max58.660.5$4.4368/s

75GPT-558.455.0$10.0076/s

76GLM-5.3-Flash57.763.1$0.078122/s

77Apodex 1.157.458.9——

78Claude Sonnet 4.657.257.0$15.0044/s

78GPT-5.157.255.3$10.0050/s

80Inkling57.154.8$4.0585/s

81Grok 4.357.053.3$2.5061/s

81Gemini 3 Flash Preview57.057.5$3.0078/s

83Beam 1.0 501B-A23B56.560.2—345/s

84Qwen3.5 397B (Reasoning)56.355.3$3.60—

85Kimi K2.656.158.2$3.4181/s

85GPT-5 (high)56.153.6$10.0076/s

87Qwen3.5 Plus Thinking55.848.1——

87Grok 455.850.0—78/s

87Grok 4.20 (Reasoning)55.855.5——

90GPT-5.1-Codex-Max55.753.8$10.0055/s

90GPT-5.1-Codex55.752.9$10.0055/s

90GLM-5.255.760.6$6.80111/s

93Kimi K2.5 (Reasoning)55.553.0$3.00—

94Grok 4.2055.452.8$2.5072/s

95Qwen3.8-Flash-Next55.264.3—55/s

96GPT-5 (medium)54.953.0$10.0076/s

97Qwen3.7 Plus54.857.7$1.2851/s

98Claude Opus 4.554.652.0$25.0039/s

99Claude Haiku 5.554.566.3$0.50159/s

99A.X K254.551.9——

101Solar Open 254.253.3——

102Qwen3.6 Plus54.154.6$1.9544/s

103Hy3 Preview53.952.3$0.6069/s

104Qwen3.5 397B53.853.8$3.6059/s

104Apodex 1.1 Mini53.855.7Free—

106Kimi K2.553.752.2$3.3236/s

106o353.751.8$8.0065/s

106GPT-5.4 mini53.754.5$4.5095/s

109Kimi K2.5 Thinking53.552.6——

110Claude Sonnet 4.5 Thinking53.353.8——

110GLM-5 Thinking53.353.4——

112Gemini 2.5 Pro53.046.6$10.0091/s

112MAI-Thinking-153.053.5——

112Nemotron 3.5 Bolt 74B-A9B53.046.4——

115Qwen3.5-122B-A10B52.951.9$2.0870/s

116Nemotron 3.5 Super VL 120B A12B52.751.0——

116Nemotron 3.5 Super VL 120B A12B NVFP452.750.8——

118Gemini 3.1 Flash Lite Preview52.349.6$1.50151/s

118MiMo-V2.6-Flash52.364.3$0.2828/s

120MiMo-V2-Pro52.051.8——

120Inkling-Small52.055.1$1.2080/s

122GLM-551.851.6$3.2045/s

123MiniMax M2.151.749.1$1.2034/s

123GLM-5.151.756.0$4.40102/s

125MiMo-V2.5-Pro51.256.3$1.8047/s

125Qwen3.8-27B51.259.2$1.88113/s

125GLM-5-Turbo51.251.9$4.0025/s

125Qwen3 235B 250751.242.3$0.8048/s

125DeepSeek V3.2 (Thinking)51.249.4$2.19—

130Gemini 2.5 Flash (Preview)51.046.6——

131GLM-4.750.950.0$2.2028/s

132MiMo-V2.550.554.8$2.004/s

133MiniMax M2.750.251.2$0.8435/s

134Solar Pro 450.152.1$0.6044/s

135Nemotron 3 Ultra50.050.8Free—

135Gemini 3.5 Flash-Lite50.052.0$2.50130/s

137Qwen3.6-35B-A3B49.850.6$0.9065/s

138Claude Opus 4.149.447.5$75.0012/s

138Claude Sonnet 4.549.447.9$15.0045/s

140Kimi K2 Thinking49.248.2$2.5048/s

141Qwen3.5-27B49.051.3$2.6069/s

142MiniMax M2.548.950.2$1.0844/s

142Qwen3 235B 2507 (Reasoning)48.947.4$3.50108/s

144GPT-4.548.838.8$150.00—

145Muse Glimmer 30B48.753.0$1.2090/s

146Ternary Bonsai 2 27B48.652.2$0.5021/s

147Claude Opus 448.243.9——

147Qwen3.6-27B48.252.1$3.2060/s

149Ripple-65B Preview48.152.3$0.6063/s

149Grok 4.1 Fast (Reasoning)48.147.3——

149Qwen3 Max48.145.0$3.9062/s

149Claude Sonnet 4 Thinking48.146.9——

153Claude 4.1 Opus Thinking47.947.0—29/s

154Qwen3.5-35B-A3B47.449.4$1.0043/s

155Claude 4 Opus Thinking47.147.0——

155Grok 3 [Beta]47.139.3——

157DeepSeek V3.1 (Reasoning)47.043.0——

158GLM-5V-Turbo46.948.1$4.0030/s

159Step 3.7 Flash46.855.3$1.15112/s

159o146.844.5$60.0097/s

161Gemma 4 26B A4B46.744.5Free—

162Grok 4 Fast (Reasoning)46.645.3——

163GPT-5 mini46.148.4$2.0078/s

163Ling 3.0 Flash46.150.9$0.06373/s

165DeepSeek-R1-052846.044.6$2.1520/s

166Trinity-Large-Thinking45.543.5$0.80138/s

167GLM-4.545.145.7$2.2025/s

168MiMo-V2-Omni44.947.2——

169Gemma 4 31B44.647.9Free35/s

170GPT-5.4 nano44.550.9$1.2591/s

170Claude Sonnet 444.543.0$15.0049/s

172Claude 3.7 Sonnet Thinking44.141.4——

173K-EXAONE 2.043.748.8——

173Mistral Large 443.760.7$2.10113/s

175Kimi K2 Instruct43.643.2——

175Claude Haiku 4.5 Thinking43.643.5——

177Spark-X2.6-4B43.548.7——

177DeepSeek-R143.538.9$2.5018/s

179DeepSeek V3.243.244.2$0.3819/s

180Llama4 Maverick Instruct Basic42.736.0——

181Sarvam 105B42.436.7—191/s

181GPT-OSS 120B42.443.4$0.6034/s

183DeepSeek V3.141.537.9—98/s

183Kimi K241.539.6$2.3037/s

183Gemma 4 12B41.539.0——

186Mercury 2.541.446.2$0.15125/s

187Mistral Medium 3.5 128B41.250.3$7.50—

187GPT-4.141.238.0$8.0074/s

189Qwen3 235B-A22B41.138.1$1.8244/s

189Qwen3.5-4B41.144.3—20/s

191o4-mini40.844.7$4.40122/s

192ZAYA1-8B40.138.2——

193Claude 3.7 Sonnet40.037.7——

194Mistral Small 439.937.9$0.60114/s

195LongCat-Flash-Lite-Sparse39.739.3——

196K-Exaone39.642.3——

197Nemotron 3.5 Lightning 30B A3B NVFP439.537.0——

198o3-mini39.040.0$4.40173/s

198GLM-4.639.041.8$2.0030/s

198Trinity-Large-Preview39.042.2$1.00—

201Mistral Large 338.934.3$1.5061/s

201Grok Code Fast 138.941.2$1.50172/s

203North Mini Code38.738.0Free70/s

204LFM2.5-24B-A2B38.639.8$0.70280/s

204Nemotron Ultra 253B38.635.9——

206QwQ-32B38.436.3——

207Nemotron 3 Nano Omni 30B A3B38.238.6—228/s

208Grok 4.1 Fast Non Reasoning38.136.2——

209MiMo-V2-Flash38.043.4—129/s

210DeepSeek V3 032437.637.1——

211Gemini 2.5 Flash37.535.0$2.5095/s

212Claude 3.5 Sonnet37.432.2$15.00—

212Solar Pro 337.439.2$0.60114/s

212GPT-5 nano37.441.9$0.4081/s

215Exaone 4.0 32B37.232.4——

216Sarvam 30B36.830.3——

217Claude Haiku 4.536.143.2$5.0088/s

218GPT-4.1 mini36.034.5$1.6072/s

218GPT-OSS 20B36.041.7$0.13310/s

218Claude 4 Sonnet36.039.5$15.0040/s

221Gemini 2.0 Flash35.729.5——

222Command A+35.540.2$1.50161/s

222Celeris-135.533.7$0.701,555/s

224Magistral Medium35.233.7——

225Granite 4.2 30B35.041.6——

226Nemotron 3 Nano 30B34.732.9—270/s

226MiniCPM5-2B34.737.1——

228DeepSeek V334.329.9$1.1037/s

229GPT-4o34.128.7$10.0066/s

230Grok 234.029.1——

230Llama 4 Maverick34.034.4$1.0061/s

232Phi-4-reasoning33.327.1—41/s

232Granite 4.2 8B33.337.6$0.15101/s

234Grok 4.1 Fast32.533.3$0.50138/s

235Mistral Medium 331.127.1$2.0094/s

236Magistral Small31.031.4——

237Ling 2.6 Flash30.837.0—104/s

238Gemma 4 E4B30.228.2——

239Phi-429.922.4$0.1493/s

240LFM2.5-2.6B29.529.3$0.50189/s

241Llama 3.3 70B29.128.2——

242LFM3-1.7B28.841.5$0.20218/s

243Mistral Large 228.727.4$6.0070/s

244o1-mini28.233.5$4.40—

245Llama 4 Scout28.026.8$0.3063/s

246Granite 4.2 3B27.833.2——

247Claude 3 Opus27.623.3$75.00—

247Llama 3.1 70B27.624.8——

249Nova Pro27.525.6—129/s

250Gemini 2.5 Flash-Lite27.127.1$0.40140/s

251GPT-4.1 nano26.726.3$0.4056/s

251Command A26.723.9——

253Llama 3.1 405B26.327.0—29/s

254LFM2.5-8B-A1B26.030.6—335/s

255GPT-4o mini25.726.2$0.6056/s

255Mistral Small 3.125.724.2—28/s

257Gemma 3 27B24.925.3—31/s

258Claude 3.5 Haiku24.424.1——

259Mistral Medium 3.524.342.7$7.5092/s

260Devstral Small24.228.2——

261Claude 3 Haiku24.122.0$1.2573/s

262Gemini 1.5 Flash22.320.2——

263Gemma 4 E2B22.121.9——

264Llama 3.1 8B14.417.9——

265Granite-4.0-H-350M11.512.7——

266Granite-4.0-350M11.211.9——

267Granite-4.0-H-1B11.116.3——