Model Rankings.
Dashboard
Most powerful AI models · updated July 1, 2026
Ranked by Artificial Analysis Intelligence Index
1
Claude Fable 5
Anthropic
98
2
Claude Mythos 5
Anthropicgated
98
3
Claude Opus 4.8
Anthropic
91
4
GPT-5.5
OpenAI
90
5
Claude Opus 4.7
Anthropic
88
6
Claude Sonnet 5
Anthropic
86
7
GLM-5.2
Zhipu
83
8
Gemini 3.5 Flash
Google
82
9
Claude Sonnet 4.6
Anthropic
77
10
Gemini 3.1 Pro
Googlepreview
75
11
Qwen3.7 Max
Alibaba
75
12
MiniMax-M3
MiniMax
72
13
DeepSeek V4 Pro
DeepSeek
72
14
GPT-5.3 Codex
OpenAI
70
15
Grok 4.1
xAI
69
Ranked by SWE-bench Verified & Pro
1
Claude Mythos 5
Anthropicgated
96
2
Claude Fable 5
Anthropic
95
3
Claude Opus 4.8
Anthropic
89
4
Claude Opus 4.7
Anthropic
87
5
GPT-5.3 Codex
OpenAI
85
6
Claude Sonnet 5
Anthropic
84
7
GPT-5.5
OpenAI
83
8
GLM-5.2
Zhipu
80
9
Gemini 3.5 Flash
Google
79
10
DeepSeek V4 Pro
DeepSeek
78
11
Qwen3.7 Max
Alibaba
77
12
Gemini 3.1 Pro
Googlepreview
75
13
MiniMax-M3
MiniMax
72
14
Grok 4.1
xAI
70
Ranked by Arena (LMArena) human-preference Elo
1
GPT-5
OpenAI
98
2
Claude Opus 4.8
Anthropic
96
3
Claude Fable 5
Anthropic
95
4
Gemini 3.1 Pro
Googlepreview
93
5
Claude Opus 4.6
Anthropic
92
6
GPT-5.5
OpenAI
91
7
Gemini 3.5 Flash
Google
89
8
Claude Sonnet 5
Anthropic
87
9
Grok 4
xAI
85
10
DeepSeek V4 Pro
DeepSeek
82
11
GLM-5.2
Zhipu
80
12
Qwen3.7 Max
Alibaba
78
13
MiniMax-M3
MiniMax
75
14
Llama 4 Behemoth
Meta
72
Ranked by CyBench + Cybersecurity-CTFs benchmark (offensive-CTF numbers favor 4.7; overall frontier favors 4.8 — close pair)
1
Claude Mythos 5
Anthropicgated
99
2
Claude Fable 5
Anthropic
96
3
Claude Opus 4.8
Anthropic
95
4
Claude Opus 4.7
Anthropic
93
5
Claude Opus 4.6
Anthropic
91
6
GPT-5.3 Codex
OpenAI
78
7
GPT-5.5
OpenAI
72
8
Claude Sonnet 5
Anthropic
70
9
Gemini 3.5 Flash
Google
64
10
GLM-5.2
Zhipu
60
11
Claude Haiku 4.5
Anthropic
56
12
DeepSeek V4 Pro
DeepSeek
52
13
Grok 4.1
xAI
48
14
Qwen3.7 Max
Alibaba
44
Ranked 0–100 by intelligence/power · refreshed daily · bar color = provider company.