Artificial Analysis Intelligence Index

Name: Artificial Analysis Intelligence Index leaderboard
Creator: Artificial Analysis

Intelligence

The most-cited composite intelligence score: a 0–100 index combining knowledge, reasoning, math, coding, and agentic evaluations (GPQA Diamond, HLE, IFBench, SciCode, Terminal-Bench Hard, τ²-Bench, and more). Higher is better.

Official source

Leaderboard

#ModelScorePrice

1
Claude Fable 5Adaptive Reasoning, Max Effort, Opus 4.8 Fallback
59.9$20.00
2
GPT-5.6 Solmax
58.9$11.25
3
GPT-5.6 Solxhigh
57.7$11.25
4
GPT-5.6 Solhigh
55.9$11.25
5
Claude Opus 4.8Adaptive Reasoning, Max Effort
55.7$10.00
6
GPT-5.6 Terramax
55$5.63
7
GPT-5.5xhigh
54.8$11.25
8
Grok 4.5high
53.8$3.00
9
GPT-5.6 Solmedium
53.6$11.25
10
Claude Opus 4.7Adaptive Reasoning, Max Effort
53.5$10.00
11
Claude Sonnet 5Adaptive Reasoning, Max Effort
53.4$4.00
12
GPT-5.5high
53.1$11.25
13
GPT-5.6 Terraxhigh
51.6$5.63
14
GPT-5.4xhigh
51.4$5.63
15
GPT-5.6 Lunamax
51.2$2.25
16
GLM-5.2max
51.1$2.15
17
GPT-5.5medium
50.4$11.25
18
Gemini 3.5 Flashhigh
50.2$3.38
19
GPT-5.6 Sollow
49.4$11.25
20
GPT-5.6 Lunaxhigh
49.1$2.25
21
GPT-5.6 Terrahigh
49$5.63
22
Claude Sonnet 4.6Adaptive Reasoning, Max Effort
47.2$6.00
23
Gemini 3.1 Pro Preview
46.5$4.50
24
GPT-5.6 Lunahigh
46.1$2.25
25
Qwen3.7 Max
46$3.75
26
GPT-5.6 Terramedium
45.6$5.63
27
Gemini 3.5 Flashmedium
45.4$3.38
28
MiniMax-M3
44.4$0.53
29
DeepSeek V4Reasoning, Max Effort
44.3$0.54
30
Kimi K2.6
44.2$1.71
31
GPT-5.5low
43.5$11.25
32
Claude Opus 4.7Non-reasoning, High Effort
42.7$10.00
33
GPT-5.2xhigh
42.2$4.81
34
Kimi K2.7 Code
41.9$1.71
35
Claude Sonnet 5Non-reasoning, High Effort
41.7$4.00
36
GPT-5.6 SolNon-reasoning
41.2$11.25
37
DeepSeek V4Reasoning, High Effort
40.8$0.54
38
GPT-5.6 Terralow
40.5$5.63
39
GPT-5.2 Codexxhigh
40.1$4.81
40
GPT-5.4 Minixhigh
40$1.69
41
Gemini 3 Prohigh
39.6$4.50
42
GPT-5.4low
39.1$5.63
43
Kimi K2.5Reasoning
38.1$1.20
44
GPT-5.6 Lunamedium
38.1$2.25
45
GPT-5.2medium
38$4.81
46
Gemini 3 FlashReasoning
37.8$1.13
47
Grok 4.3high
37.6$1.56
48
Grok 4.3medium
36$1.56
49
Claude Sonnet 4.6Non-reasoning, High Effort
35.9$6.00
50
GPT-5.5Non-reasoning
35.4$11.25
51
Grok 4.3low
35.4$1.56
52
Gemini 3.5 Flashminimal
34.9$3.38
53
Kimi K2.6Non-reasoning
34.6$1.71
54
Claude Sonnet 4.6Non-reasoning, Low Effort
34.3$6.00
55
GPT-5.6 TerraNon-reasoning
34$5.63
56
GPT-5.5May 2026
33.5$11.25
57
GPT-5.6 Lunalow
33.3$2.25
58
Gemini 3 Prolow
33.1$4.50
59
DeepSeek V4Non-reasoning
31.2$0.54
60
GPT-5.4 Minimedium
29.8$1.69
61
Claude Haiku 4.5Reasoning
29.6$2.00
62
Kimi K2.5Non-reasoning
29.4$1.20
63
GPT-5.5June 2026
28.9$11.25
64
GPT-5.4Non-reasoning
27.7$5.63
65
Gemini 3 FlashNon-reasoning
27.4$1.13
66
GPT-5.6 LunaNon-reasoning
26.6$2.25
67
GPT-5.2Non-reasoning
26$4.81
68
Grok 4.3Non-reasoning
24.8$1.56
69
Claude Haiku 4.5Non-reasoning
23.7$2.00
70
GPT-5.4 MiniNon-Reasoning
16.6$1.69
71
Mistral Large 3
15.9$0.75

Sources:

AA Data API (models endpoint)Artificial Analysis Intelligence Index

Used in rankings:

Best cheap AI models Best Chinese AI models Best Qwen models

Share:

Details:

Category
Intelligence
Created by
Artificial Analysis
Models tested
28
Configs tested
71
Leader
Claude Fable 5
Top score
59.9

Updated July 2026