English Français العربية

Most capable

Artificial Analysis Intelligence Index, current language models.

History since Oct 23, 2026. Movement appears once a week of history is recorded.

RankModelValueRank, last 30 days
1 Claude Opus 5.5Anthropic 57.6
2 Claude Sonnet 5.5Anthropic 56.0
3 Muse Wonder FullMeta 55.8
4 Claude Fable 5.1Anthropic 53.4
5 GPT-6 AstraOpenAI 52.7
6 Gemini 4 ArgonGoogle 52.6
7 Muse WonderMeta 52.4
8 GPT-6.1 SolOpenAI 51.8
9 Claude Opus 5Anthropic 50.8
10 Claude Fable 5Anthropic 49.6
11 Muse Spark 1.3Meta 48.1
12 GPT-6 SolOpenAI 47.6
13 GPT-5.6 SolOpenAI 47.0
14 Grok 4.7xAI 46.4
15 Qwen3.8-MaxAlibaba Cloud (Qwen) 45.4
16 DeepSeek V4.1 FlashDeepSeek 45.1
17 Z.ai GLM 5.3 (on Mistral)Mistral AI 44.8
18 Grok 4.6xAI 44.3
19 Kimi K3Moonshot AI 43.6
20 Claude Haiku 5.5Anthropic 43.4
21 GPT-5.6 TerraOpenAI 42.1
22 Claude Opus 4.8Anthropic 41.8
23 Gemini 3.8 FlashGoogle 40.9
24 Muse Spark 1.2Meta 39.6
25 Muse Spark OpenMeta 38.9
26 Grok 4.5xAI 38.8
27 Mistral Large 4Mistral AI 38.4
28 GPT-5.5OpenAI 38.4
29 Claude Sonnet 5Anthropic 38.2
30 GPT-6 LunaOpenAI 38.1
31 GPT-5.6 LunaOpenAI 37.3
32 DeepSeek V4 ProDeepSeek 36.0
33 Gemini 3.6 FlashGoogle 34.0
34 Muse Spark 1.1Meta 33.7
35 Gemini 3.1 ProGoogle 29.7
36 Kimi K2.6Moonshot AI 27.0
37 Kimi K2.7 CodeMoonshot AI 25.8
38 Qwen3.7-PlusAlibaba Cloud (Qwen) 25.2
39 Grok 4.3xAI 24.9
40 GPT-5.4 MiniOpenAI 24.1
41 Gemini 3.5 Flash-LiteGoogle 22.2
42 GPT-5.4 nanoOpenAI 20.7
43 Muse GlimmerMeta 17.5
44 GPT-5 MiniOpenAI 16.8
45 Gemini 2.5 ProGoogle 16.1
46 Mistral Medium 3.5Mistral AI 14.2
47 Command A+Cohere 13.1
48 gpt-oss-120bOpenAI 11.6
49 Mistral Small 4Mistral AI 11.3
50 North Mini CodeCohere 9.9

Rankings are recorded once a day from the catalogue: benchmark scores and speeds from Artificial Analysis, prices from each provider. Movement compares today with the closest day at least a week ago.