最佳长上下文 AI 模型

107 个模型 · 指数。上下文 1M token 及以上。方法

上榜模型

2026年11月

#模型指数输出 / 1M速度

01Claude Fable 5.285.6$40.0064/s

02Gemini 4 Carbon82.7$10.0031/s

03Claude Opus 5.578.6$20.0076/s

04Gemini 4 Argon78.5$10.0028/s

05GPT-6 Astra78.0$50.0034/s

05GPT-6.1 Sol78.0$10.0034/s

07Muse Wonder Full77.3——

08Claude Sonnet 5.577.2$10.00107/s

09GPT-6.1 Terra77.1$5.0046/s

10Claude Fable 5.176.2$50.0069/s

11Kimi K3.1 Pro75.7$10.0043/s

12GPT-5.5 Pro74.8$180.0019/s

13Claude Fable 574.0$50.0056/s

14Claude Opus 573.8$25.0048/s

15GPT-5.4 Pro73.3$180.0021/s

16Muse Wonder73.1$20.0080/s

17GPT-5.6 Sol73.0$22.0064/s

18GPT-6 Sol72.8$10.0061/s

19Hy472.0$1.2552/s

20Claude Mythos 570.8$50.00—

21MiMo-V2.6-Pro70.2$0.8731/s

22Muse Spark 1.370.1$4.2576/s

23MiniMax M3.1 Flash69.3$0.60100/s

24GPT-5.569.1$30.0052/s

25GPT-5.6 Terra68.9$12.0066/s

26DeepSeek V4.1 Flash 101768.5$0.40261/s

26Gemini 3.7 Flash68.5$3.7585/s

28Step 5 Preview68.4$2.70105/s

29Gemini 3.8 Flash67.8$4.1368/s

30Claude Opus 4.867.7$25.0069/s

31Ling 3.1 Flash67.2$0.13581/s

32Kimi K366.9$16.0044/s

33Hy4 preview66.4$2.5067/s

33Inkling 2 Small66.4$1.00—

35Sakana Fugu-Ultra66.3——

35Claude Haiku 5.566.3$0.50159/s

37GLM-5.365.8$1.4086/s

38Qwen3.8 Max65.6—41/s

39GPT-6 Luna65.5$0.5066/s

39DeepSeek V4 Pro 081365.5$4.4442/s

39GPT-5.6 Luna65.5$1.2051/s

42Muse Spark Open65.1$4.2586/s

43Muse Spark 1.264.9$4.25123/s

44Muse Ember64.8$0.50158/s

44GPT-5.464.8$15.0038/s

46Muse Spark 1.164.6$4.25152/s

47MiMo-V2.6-Flash64.3$0.2828/s

48Claude Sonnet 564.2$10.0075/s

48Gemini 3.5 Flash64.2$9.00137/s

50Sakana Fugu63.9——

51Gemini 3.6 Flash63.7$3.75121/s

52Qwen3.8 Max (0902)63.4$6.0033/s

53Claude Opus 4.7 (Adaptive)63.2$25.00—

54GLM-5.3-Flash63.1$0.078122/s

55Qwen3.8-Omni-Flash62.7$0.4741/s

56Gemini 3.1 Pro62.3$12.00151/s

57Claude Opus 4.762.0$25.0047/s

58DeepSeek V4 Flash 073161.9$0.3061/s

59Gemini 3.1 Pro Preview61.5$12.0077/s

60Mistral Large 460.7$2.10113/s

61GLM-5.260.6$6.80111/s

62Qwen3.7 Max60.5$4.4368/s

63DeepSeek V4 Flash Vision Exp60.4$1.3274/s

64Beam 1.0 501B-A23B60.2—345/s

65Claude Opus 4.6 (Adaptive)60.0——

66Beam 1.0 501B-A23B FP859.2——

66Qwen3.8-27B59.2$1.88113/s

68Beam 1.0 501B-A23B NVFP458.5——

69Claude Opus 4.658.0$25.0036/s

70Gemini 3 Pro57.9$12.00109/s

71Qwen3.7 Plus57.7$1.2851/s

72DeepSeek V4 Pro 042357.6$1.7889/s

73Gemini 3 Flash Preview57.5$3.0078/s

74Claude Sonnet 4.657.0$15.0044/s

75MiMo-V2.5-Pro56.3$1.8047/s

76MiniMax M355.7$1.2079/s

77Gemini 3 Flash55.0$3.00198/s

78MiMo-V2.554.8$2.004/s

79Qwen3.6 Plus54.6$1.9544/s

80Grok 4.353.3$2.5061/s

80Solar Open 253.3——

82Grok 4.2052.8$2.5072/s

83Mercury 352.3$0.35295/s

83Ripple-65B Preview52.3$0.6063/s

85Gemini 3.5 Flash-Lite52.0$2.50130/s

86Gemini 3.1 Flash-Lite51.9$1.50134/s

87MiMo-V2-Pro51.8——

88Nemotron 3.5 Super VL 120B A12B51.0——

89Nemotron 3.5 Super VL 120B A12B NVFP450.8——

89Nemotron 3 Ultra50.8Free—

91Gemini 3.1 Flash Lite Preview49.6$1.50151/s

92Spark-X2.6-4B48.7——

93Claude Sonnet 4.547.9$15.0045/s

94Grok 4.1 Fast (Reasoning)47.3——

95Gemini 2.5 Pro46.6$10.0091/s

96Grok 4 Fast (Reasoning)45.3——

97LongCat-Flash-Lite-Sparse39.3——

98GPT-4.138.0$8.0074/s

99Nemotron 3.5 Lightning 30B A3B NVFP437.0——

100Gemini 2.5 Flash35.0$2.5095/s

101GPT-4.1 mini34.5$1.6072/s

102Grok 4.1 Fast33.3$0.50138/s

103Gemini 2.0 Flash29.5——

104Gemini 2.5 Flash-Lite27.1$0.40140/s

105Llama 4 Scout26.8$0.3063/s

106GPT-4.1 nano26.3$0.4056/s

107Gemini 1.5 Flash20.2——