Skip to main content

Benchmark profile

Artificial Analysis Intelligence Index

A display-only intelligence index published by Artificial Analysis that aggregates provider-reported and benchmark-derived signals into a single model-level score.

Data verified

The public Artificial Analysis Intelligence Index snapshot ranks Claude Fable 5 first at 59.9%, ahead of GPT-5.6 Sol (58.9%) and Kimi K3 (57.1%) among 166 tested models. We mirror the table as display-only evidence; it does not affect overall rankings.

Benchmark score on Artificial Analysis Intelligence Index — July 23, 2026

BenchLM mirrors the published score view for Artificial Analysis Intelligence Index. Claude Fable 5 leads the public snapshot at 59.9% , followed by GPT-5.6 Sol (58.9%) and Kimi K3 (57.1%). BenchLM does not use these results to rank models overall.

166 modelsKnowledgeCurrentDisplay onlyUpdated July 23, 2026

Benchmark score table (166 models)

Score
1
Claude Fable 5Anthropic · Closed
59.9%
2
GPT-5.6 SolOpenAI · Closed
58.9%
3
Kimi K3Moonshot AI · Closed
57.1%
4
Claude Opus 4.8Anthropic · Closed
55.7%
5
GPT-5.6 TerraOpenAI · Closed
55.0%
6
GPT-5.5OpenAI · Closed
54.8%
7
Grok 4.5xAI · Closed
53.8%
8
Claude Opus 4.7 (Adaptive)Anthropic · Closed
53.5%
9
Claude Sonnet 5Anthropic · Closed
53.4%
10
GPT-5.4OpenAI · Closed
51.4%
11
GPT-5.6 LunaOpenAI · Closed
51.2%
12
GLM-5.2Z.AI · Open weight
51.1%
13
Muse Spark 1.1Meta · Closed
50.6%
14
Gemini 3.5 FlashGoogle · Closed
50.2%
15
Gemini 3.6 FlashGoogle · Closed
50.1%
16
Gemini 3.1 ProGoogle · Closed
46.5%
17
Qwen3.7 MaxAlibaba · Closed
46.0%
18
MiniMax M3MiniMax · Open weight
44.4%
19
DeepSeek V4 Pro (Max)DeepSeek · Open weight
44.3%
20
GPT-5.3 CodexOpenAI · Closed
44.3%
21
GPT-5.3-Codex-SparkOpenAI · Closed
44.3%
22
Kimi K2.6Moonshot AI · Open weight
44.2%
23
Claude Opus 4.6 (Adaptive)Anthropic · Closed
43.7%
24
DeepSeek V4 Pro (High)DeepSeek · Open weight
43.1%
25
Muse SparkMeta · Closed
43.1%
26
Claude Opus 4.7Anthropic · Closed
42.7%
27
MiMo-V2.5-ProXiaomi · Closed
42.2%
28
GPT-5.2OpenAI · Closed
42.2%
29
Kimi K2.7 CodeMoonshot AI · Open weight
42.0%
30
Hy3 PreviewTencent · Open weight
41.2%
31
Hy3Tencent · Open weight
41.2%
32
Claude Opus 4.5 ThinkingAnthropic · Closed
40.8%
33
InklingThinking Machines Lab · Open weight
40.7%
34
MiMo-V2-ProXiaomi · Closed
40.3%
35
DeepSeek V4 Flash (Max)DeepSeek · Open weight
40.3%
36
GLM-5.1Z.AI · Open weight
40.2%
37
GPT-5.2-CodexOpenAI · Closed
40.1%
38
Qwen 3.6 Max (preview)Alibaba · Closed
40.0%
39
GPT-5.4 miniOpenAI · Closed
40.0%
40
Qwen3.6 PlusAlibaba · Closed
39.6%
41
Gemini 3 ProGoogle · Closed
39.5%
42
GLM-5Z.AI · Open weight
39.5%
43
Qwen3.7 PlusAlibaba · Closed
39.0%
44
GPT-5.4 nanoOpenAI · Closed
38.2%
45
MiniMax M2.7MiniMax · Open weight
38.1%
46
GLM-5-TurboZ.AI · Closed
38.1%
47
Claude Opus 4.6Anthropic · Closed
37.8%
48
Nemotron 3 UltraNVIDIA · Open weight
37.8%
49
Grok 4.3xAI · Closed
37.6%
50
DeepSeek V4 Flash (High)DeepSeek · Open weight
37.5%
51
Qwen3.6-27BAlibaba · Open weight
37.0%
52
GPT-5.1OpenAI · Closed
36.9%
53
Gemini 3.5 Flash-LiteGoogle · Closed
36.5%
54
Claude Sonnet 4.6Anthropic · Closed
35.9%
55
Kimi K2.5Moonshot AI · Open weight
35.4%
56
Kimi K2.5 (Reasoning)Moonshot AI · Closed
35.4%
57
MiMo-V2-OmniXiaomi · Closed
35.0%
58
GPT-5.1-Codex-MaxOpenAI · Closed
34.7%
59
GPT-5.1-CodexOpenAI · Closed
34.7%
60
Claude Opus 4.5Anthropic · Closed
34.7%
61
GPT-5 (high)OpenAI · Closed
34.7%
62
GLM-5V-TurboZ.AI · Closed
34.5%
63
Qwen3.5-27BAlibaba · Open weight
33.8%
64
GPT-5 (medium)OpenAI · Closed
33.7%
65
Claude 4.1 Opus ThinkingAnthropic · Closed
33.7%
66
GLM-4.7Z.AI · Open weight
33.7%
67
Qwen3.5 397BAlibaba · Open weight
33.7%
68
Qwen3.5 397B (Reasoning)Alibaba · Open weight
33.7%
69
MiniMax M2.5MiniMax · Closed
33.6%
70
Grok 4xAI · Closed
33.3%
71
o3-proOpenAI · Closed
32.5%
72
Qwen3.5-122B-A10BAlibaba · Open weight
32.3%
73
Qwen3.6-35B-A3BAlibaba · Open weight
31.6%
74
30.6%
75
o3OpenAI · Closed
30.4%
76
Step 3.7 FlashStepFun · Open weight
30.3%
77
Mistral Medium 3.5 128BMistral · Open weight
29.9%
78
Gemma 4 31BGoogle · Open weight
29.4%
79
Qwen3.5-35B-A3BAlibaba · Open weight
29.3%
80
Claude 4.1 OpusAnthropic · Closed
28.2%
81
Gemini 3 FlashGoogle · Closed
27.4%
82
27.4%
83
Step 3.5 FlashStepFun · Open weight
26.0%
84
Gemini 2.5 ProGoogle · Closed
25.8%
85
Gemma 4 26B A4BGoogle · Open weight
25.7%
86
Claude 4 SonnetAnthropic · Closed
25.5%
87
Nemotron 3 Super 120B A12BNVIDIA · Open weight
25.4%
88
GPT-5 miniOpenAI · Closed
25.3%
89
Gemini 3.1 Flash-LiteGoogle · Closed
25.0%
90
K-ExaoneLG AI Research · Closed
24.7%
91
MiMo-V2-FlashXiaomi · Open weight
24.7%
92
DeepSeek V3.2DeepSeek · Open weight
24.7%
93
Trinity-Large-PreviewArcee AI · Open weight
24.5%
94
Trinity-Large-ThinkingArcee AI · Open weight
24.5%
95
Qwen3 MaxAlibaba · Closed
24.0%
96
GPT-OSS 120BOpenAI · Open weight
23.8%
97
o1OpenAI · Closed
23.4%
98
GLM-4.6Z.AI · Open weight
23.0%
99
GLM-4.7-FlashZ.AI · Open weight
22.9%
100
Command A+Cohere · Open weight
22.5%
101
Gemma 4 12BGoogle · Open weight
22.0%
102
Grok Code Fast 1xAI · Closed
21.6%
103
Mercury 2Inception · Closed
21.4%
104
DeepSeek V3.1DeepSeek · Open weight
21.1%
105
DeepSeek V3.1 (Reasoning)DeepSeek · Open weight
20.7%
106
DeepSeek-R1DeepSeek · Open weight
20.1%
107
GPT-5 nanoOpenAI · Closed
19.9%
108
Mistral Small 4Mistral · Open weight
19.6%
109
Mistral Small 4 (Reasoning)Mistral · Open weight
19.6%
110
Kimi K2Moonshot AI · Closed
19.4%
111
GPT-4.1OpenAI · Closed
19.4%
112
o3-miniOpenAI · Closed
19.0%
113
o1-proOpenAI · Closed
18.9%
114
MiniMax M1 80kMiniMax · Closed
17.7%
115
o1-previewOpenAI · Closed
17.0%
116
Grok 4.1 FastxAI · Closed
16.9%
117
GLM-4.5-AirZ.AI · Closed
16.5%
118
Mistral Large 3Mistral · Closed
15.9%
119
Nemotron 3 Nano Omni 30B A3BNVIDIA · Open weight
14.9%
120
GPT-OSS 20BOpenAI · Open weight
14.9%
121
GPT-4.1 miniOpenAI · Closed
14.8%
122
Llama 4 MaverickMeta · Open weight
14.3%
123
Nemotron 3 Nano 30BNVIDIA · Open weight
14.2%
124
DeepSeek V3DeepSeek · Open weight
14.2%
125
Gemini 2.5 FlashGoogle · Closed
14.1%
126
Ling 2.6 FlashInclusionAI · Open weight
14.1%
127
Gemma 4 E4BGoogle · Open weight
12.5%
128
Mistral Medium 3Mistral · Closed
12.5%
129
Sarvam 105BSarvam · Open weight
11.9%
130
Claude 3 OpusAnthropic · Closed
11.8%
131
GPT-4oOpenAI · Closed
11.2%
132
Ministral 3 14B (Reasoning)Mistral · Open weight
11.1%
133
Ministral 3 14BMistral · Open weight
11.1%
134
DeepSeek R1 Distill Qwen 32BDeepSeek · Open weight
11.0%
135
Llama 4 ScoutMeta · Open weight
10.0%
136
Gemini 1.5 ProGoogle · Closed
10.0%
137
GPT-4.1 nanoOpenAI · Closed
9.6%
138
Gemma 4 E2BGoogle · Open weight
9.3%
139
Mistral Large 2Mistral · Closed
9.2%
140
Nemotron Ultra 253BNVIDIA · Open weight
9.1%
141
Ministral 3 8B (Reasoning)Mistral · Open weight
9.0%
142
Ministral 3 8BMistral · Open weight
9.0%
143
Llama 3.1 405BMeta · Open weight
8.5%
144
LFM2.5-8B-A1BLiquidAI · Open weight
8.3%
145
GPT-4 TurboOpenAI · Closed
7.9%
146
Solar Pro 2Upstage · Closed
7.8%
147
Nova ProAmazon · Closed
7.7%
148
Gemma 3 27BGoogle · Open weight
7.4%
149
Qwen2.5 Coder 32B InstructAlibaba · Open weight
7.1%
150
GPT-4o miniOpenAI · Closed
6.9%
151
Ministral 3 3B (Reasoning)Mistral · Open weight
6.8%
152
Ministral 3 3BMistral · Open weight
6.8%
153
Sarvam 30BSarvam · Open weight
6.6%
154
Exaone 4.0 32BLG AI Research · Open weight
6.0%
155
LFM2-24B-A2BLiquidAI · Closed
5.0%
156
Phi-4Microsoft · Open weight
4.9%
157
Claude 3 HaikuAnthropic · Closed
3.9%
158
Gemini 1.0 ProGoogle · Closed
3.1%
159
Exaone 4.0 1.2BLG AI Research · Open weight
2.8%
160
LFM2.5-1.2B-ThinkingLiquidAI · Closed
2.8%
161
LFM2.5-1.2B-InstructLiquidAI · Closed
2.7%
162
Granite-4.0-H-1BIBM · Open weight
2.7%
163
Granite-4.0-1BIBM · Open weight
2.1%
164
LFM2.5-VL-1.6B-ExtractLiquidAI · Open weight
1.0%
165
Granite-4.0-350MIBM · Open weight
1.0%
166
Granite-4.0-H-350MIBM · Open weight
1.0%

The published Artificial Analysis Intelligence Index snapshot places Claude Fable 5 first at 59.9%. The third row is 2.8 points behind. The broader top-10 range is 8.5 points, so many of the published results sit in a relatively narrow band.

166 models have been evaluated on Artificial Analysis Intelligence Index. The benchmark falls in the Knowledge category. This category carries a 12% weight in BenchLM.ai's overall scoring system. Artificial Analysis Intelligence Index is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About Artificial Analysis Intelligence Index

Year

2026

Tasks

Cross-benchmark intelligence index

Format

Aggregated model score

Difficulty

Display-only external reference

BenchLM tracks Artificial Analysis as a display-only external reference rather than a weighted benchmark. It is useful as a market snapshot, but it is not a benchmark-native row with a single public task set, scoring harness, or exact-source methodology aligned to BenchLM's core benchmark pages.

BenchLM freshness & provenance

Version

Artificial Analysis Intelligence Index 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

FAQ

What does Artificial Analysis Intelligence Index measure?

A display-only intelligence index published by Artificial Analysis that aggregates provider-reported and benchmark-derived signals into a single model-level score.

Which model scores highest on Artificial Analysis Intelligence Index?

Claude Fable 5 by Anthropic currently leads with a score of 59.9% on Artificial Analysis Intelligence Index.

How many models are evaluated on Artificial Analysis Intelligence Index?

166 AI models have been evaluated on Artificial Analysis Intelligence Index on BenchLM.

Last updated: July 23, 2026 · BenchLM version Artificial Analysis Intelligence Index 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.