코딩 AI 순위

Q1몇 시간 걸리는 개발 일도 컴퓨터를 직접 다루며 혼자 끝낼까?Terminal-Bench 4.0·Claude Sonnet 5.5🥇Claude Sonnet 5.5🥇Q2대중들이 가장 선호하는 코딩 답변을 하는 AI는?Arena Coding·Gemini 4 Argon🥇Gemini 4 Argon🥇Q3코드를 실행해 보며 고쳐 과제를 끝낼 수 있을까?LiveBench Agentic Coding·DeepSeek V4.1 Flash🥇DeepSeek V4.1 Flash🥇

그 외 평가 순위(16개)

이 평가 지표에 대해 상세히 알아보기

1. 🥇AnthropicClaude Sonnet 5.5추론 강도: MaxQ163.6%Q21536Q339.3
2. 🥈AnthropicClaude Opus 5.5추론 강도: MaxQ159.6%Q21539Q365.3
3. 🥈OpenAIGPT-6 Astra추론 강도: Extra HighQ159.6%Q21543Q357.3
4. #4GoogleGemini 4 Argon추론 강도: HighQ157.1%Q21560Q3—
5. #5OpenAIGPT-6.1 Sol추론 강도: MaxQ156.1%Q21542Q356.8
6. #6AnthropicClaude Fable 5.1추론 강도: Extra HighQ155.1%Q21531Q366.1
7. #7AnthropicClaude Opus 5추론 강도: MaxQ149.0%Q21530Q365.2
8. #8OpenAIGPT-6 Sol추론 강도: MaxQ143.9%Q21521Q352.9
9. #9AnthropicClaude Fable 5추론 강도: MaxQ142.4%Q21552Q362.2
10. #10Z.aiGLM 5.3추론 강도: MaxQ141.9%Q21521Q360.9
11. #11OpenAIGPT-5.6 Sol추론 강도: MaxQ139.9%Q21534Q356.2
12. #12OpenAIGPT-5.6 Terra추론 강도: MaxQ135.4%Q21515Q355.0
13. #13XiaomiMiMo-V2.6-ProQ134.8%Q21540Q3—
14. #14MetaMuse Spark 1.3추론 강도: MaxQ133.3%Q21539Q364.1
15. #14StepfunStep-5Q133.3%Q21511Q3—
16. #16Z.aiGLM 5.3 FlashQ132.8%Q21523Q356.8
17. #17DeepSeekDeepSeek V4.1 Flash추론 강도: MaxQ126.8%Q21528Q377.3
18. #17Mistral AIMistral Large 4Q126.8%Q2—Q3—
19. #19xAIGrok 4.7추론 강도: Extra HighQ125.8%Q21488Q354.0
20. #20AlibabaQwen3.8 FlashQ125.3%Q2—Q361.6
21. #21XiaomiMiMo-V2.6-FlashQ122.7%Q21521Q3—
22. #22AnthropicClaude Opus 4.8추론 강도: MaxQ121.7%Q21530Q350.5
23. #23xAIGrok 4.6추론 강도: HighQ121.2%Q21507Q357.0
24. #24GoogleGemini 3.8 Flash추론 강도: HighQ119.7%Q21530Q354.2
25. #25OpenAIGPT-5.5추론 강도: Extra HighQ114.6%Q21511Q354.0
26. #26AnthropicClaude Sonnet 5추론 강도: MaxQ114.1%Q21519Q359.4
27. #26DeepSeekDeepSeek V4 Pro추론 강도: MaxQ114.1%Q21500Q342.6
28. #28GoogleGemini 3.7 Flash추론 강도: HighQ113.6%Q21518Q358.3
29. #29OpenAIGPT-6 Luna추론 강도: MaxQ112.6%Q21512Q351.2
30. #29Moonshot AIKimi K3추론 강도: MaxQ112.6%Q21541Q362.2
31. #31DeepSeekDeepSeek V4 Flash추론 강도: MaxQ112.1%Q21484Q337.6
32. #32OpenAIGPT-5.6 Luna추론 강도: MaxQ111.6%Q21501Q348.4
33. #33xAIGrok 4.5추론 강도: HighQ110.6%Q21514Q356.5
34. #34GoogleGemini 3.6 Flash추론 강도: HighQ17.1%Q21519Q343.4
35. #34MetaMuse Spark 1.2추론 강도: Extra HighQ17.1%Q21531Q357.6
36. #36GoogleGemini 3.5 Flash추론 강도: HighQ16.6%Q21506Q349.0
37. #37MetaMuse Spark 1.1추론 강도: Extra HighQ16.1%Q21533Q358.5
38. #38GoogleGemini 3.1 ProQ14.0%Q21522Q344.1
39. #39AnthropicClaude Sonnet 4.6추론 강도: MaxQ13.0%Q21529Q342.6
40. #40Z.aiGLM-5.1추론 강도: ThinkingQ12.0%Q21513Q3—
41. #40OpenAIGPT-5.4 Mini추론 강도: Extra HighQ12.0%Q21495Q341.7
42. #40MiniMaxMiniMax M3Q12.0%Q21494Q340.7
43. #43AlibabaQwen3.7 MaxQ11.5%Q21524Q343.6
44. #44GoogleGemini 3.5 Flash-LiteQ11.0%Q21501Q345.3
45. #44Z.aiGLM 5.2추론 강도: MaxQ11.0%Q21512Q351.8
46. #44Moonshot AIKimi K2.7 CodeQ11.0%Q2—Q345.7
47. #44AlibabaQwen3.7 PlusQ11.0%Q21501Q3—
48. #44UpstageSolar Mini 4Q11.0%Q2—Q3—
49. #49GoogleGemini 3.1 Flash LiteQ10.5%Q21458Q3—
50. #49OpenAIGPT-5.4 Nano추론 강도: Extra HighQ10.5%Q21460Q346.8
51. #49TencentHy3Q10.5%Q21453Q3—
52. #49Moonshot AIKimi K2.6추론 강도: ThinkingQ10.5%Q21516Q346.9
53. #49NVIDIANemotron 3 Ultra추론 강도: ThinkingQ10.5%Q21474Q338.7
54. #49CohereNorth Mini CodeQ10.5%Q2—Q3—
55. #49Ring AIRing-2.6-1TQ10.5%Q2—Q3—
56. #49UpstageSolar Pro 4Q10.5%Q21463Q3—
57. #49Arcee AITrinity Large ThinkingQ10.5%Q21413Q3—
58. #58SK TelecomA.X K2Q10.0%Q2—Q3—
59. #58AnthropicClaude Haiku 4.5추론 강도: ThinkingQ10.0%Q21481Q3—
60. #58AnthropicClaude Sonnet 4.5추론 강도: ThinkingQ10.0%Q21512Q3—
61. #58GoogleGemini 2.5 ProQ10.0%Q21465Q3—
62. #58GoogleGemma 4 31B추론 강도: ThinkingQ10.0%Q21502Q3—
63. #58OpenAIGPT OSS 120B추론 강도: HighQ10.0%Q21390Q3—
64. #58OpenAIGPT-5 Mini추론 강도: HighQ10.0%Q21431Q3—
65. #58xAIGrok 4.3추론 강도: HighQ10.0%Q21488Q318.5
66. #58MetaLlama 4 MaverickQ10.0%Q21373Q3—
67. #58MetaLlama 4 ScoutQ10.0%Q21362Q3—
68. #58MeituanLongCat 2.0Q10.0%Q2—Q3—
69. #58XiaomiMiMo V2.5Q10.0%Q21491Q3—
70. #58XiaomiMiMo V2.5 Pro추론 강도: ThinkingQ10.0%Q21522Q3—
71. #58MiniMaxMiniMax M2.7Q10.0%Q21480Q3—
72. #58Mistral AIMistral Medium 3.5Q10.0%Q21479Q3—
73. #58Mistral AIMistral Small 4추론 강도: ThinkingQ10.0%Q2—Q3—
74. #58NVIDIANemotron 3 Nano 30B A3B추론 강도: ThinkingQ10.0%Q21361Q3—
75. #58NVIDIANemotron 3 Super추론 강도: ThinkingQ10.0%Q21407Q3—
76. #58AlibabaQwen3.5 397B A17B추론 강도: ThinkingQ10.0%Q21491Q3—
77. #58AlibabaQwen3.6 35B A3B추론 강도: ThinkingQ10.0%Q2—Q3—
78. #58UpstageSolar Pro 3Q10.0%Q2—Q3—