llm.ing

AA Agentic Index

run by Artificial Analysisunit: indexDisplay with attribution; no redistribution78 modelssource ↗
AA Intelligence IndexAA Coding IndexAA Agentic IndexMedian output speedMedian time to first token
#ModelScoreObserved
1Claude Fable 5.1
57.9
independent
Sep 20, 2026
2Claude Opus 5
56.5
mirror
Sep 20, 2026
3Qwen3.8 Max
56.0
mirror
Sep 20, 2026
4GLM-5.3
53.1
mirror
Sep 20, 2026
5Grok 4.6
53.0
mirror
Sep 20, 2026
6GPT-6 Astra
51.0
independent
Sep 20, 2026
7GLM-5.3-Flash
50.9
independent
Sep 20, 2026
8Claude Fable 5
50.7
mirror
Sep 20, 2026
9GPT-5.6 Sol
50.2
mirror
Sep 20, 2026
10Kimi K3
50.0
mirror
Sep 20, 2026
11GPT-5.4
44.2
mirror
Aug 7, 2026
12Claude Sonnet 5
43.6
independent
Sep 20, 2026
13GPT-5.6 Terra
43.2
mirror
Sep 20, 2026
13Muse Spark 1.2
43.2
independent
Sep 20, 2026
15GPT-5.6 Luna
42.1
independent
Sep 20, 2026
16Claude Opus 4.8
41.9
mirror
Sep 20, 2026
17DeepSeek V4 Pro
41.3
mirror
Sep 20, 2026
18Grok 4.5
41.2
mirror
Sep 20, 2026
19DeepSeek V4 Flash
41.0
mirror
Sep 20, 2026
20Gemini 3.8 Flash
40.2
mirror
Sep 20, 2026
21Claude Opus 4.7
38.6
mirror
Sep 20, 2026
22GLM-5.2
38.4
mirror
Sep 20, 2026
23GPT-5.5
36.4
independent
Sep 20, 2026
24Gemini 3.7 Flash
35.2
mirror
Sep 20, 2026
25Claude Sonnet 4.6
31.8
mirror
Sep 20, 2026
26MiniMax-M3
29.5
independent
Sep 20, 2026
27Gemini 3.6 Flash
29.0
mirror
Sep 20, 2026
27Qwen3.6 Plus
29.0
mirror
Aug 7, 2026
29Grok Build 0.1
28.9
independent
Aug 7, 2026
30GPT-5
26.5
independent
Aug 7, 2026
31GLM-4.7
26.2
mirror
Aug 7, 2026
32Gemini 3.5 Flash
26.0
independent
Sep 20, 2026
33MiniMax-M2.7
25.9
independent
Aug 7, 2026
34Muse Spark 1.1
25.8
mirror
Sep 20, 2026
35GLM-5.1
23.9
mirror
Sep 20, 2026
36Qwen3.6 27Bnon-reasoning
23.6
independent
Aug 7, 2026
37Qwen3.7 Max
22.5
mirror
Sep 20, 2026
38Kimi K2.5
21.7
mirror
Jul 23, 2026
38Step 3.7 Flash
21.7
mirror
Aug 7, 2026
40GPT-5.1
21.6
independent
Aug 7, 2026
41Kimi K2.7 Code
21.0
mirror
Sep 20, 2026
42Kimi K2.6
20.5
mirror
Sep 20, 2026
43Nemotron 3 Ultra 550B A55B
20.1
mirror
Sep 20, 2026
44GLM-4.6
18.6
mirror
Aug 7, 2026
45DeepSeek V3.2
18.3
mirror
Jul 23, 2026
46GPT-5.4 mini
17.9
mirror
Sep 20, 2026
47Qwen3.7 Plus
17.5
mirror
Sep 20, 2026
48Qwen3.5 122B-A10Bnon-reasoning
16.2
independent
Aug 7, 2026
49GPT-5.4 nano
16.0
mirror
Sep 20, 2026
50Claude Sonnet 4.5 (latest)
15.8
mirror
Sep 20, 2026
51Grok 4.3
15.5
mirror
Sep 20, 2026
52Gemini 3.5 Flash Lite
14.3
independent
Sep 20, 2026
53Qwen3.6 35B-A3B
13.1
mirror
Sep 20, 2026
54Qwen3.5 35B-A3B
11.8
mirror
Jul 23, 2026
55Gemma 4 26B A4B IT
11.0
mirror
Jul 23, 2026
56Qwen3.5 397B-A17B
8.3
independent
Sep 20, 2026
57Gemini 3.1 Pro Preview
8.2
mirror
Sep 20, 2026
58Claude Haiku 4.5 (latest)
8.0
mirror
Sep 20, 2026
59GPT-5 Mini
6.8
independent
Sep 20, 2026
60Gemma 4 31B IT
4.2
mirror
Sep 20, 2026
61Devstral 2
2.3
mirror
Sep 20, 2026
62Qwen3-Next 80B-A3B (Thinking)thinking
2.1
mirror
Jul 23, 2026
63GPT-4.1 mini
1.8
mirror
Aug 7, 2026
63Kimi K2 Thinkingthinking
1.8
mirror
Jul 23, 2026
65Nemotron 3 Super
1.7
mirror
Sep 20, 2026
66Gemini 3.1 Flash Lite Preview
1.6
mirror
Sep 20, 2026
66Gemini 2.5 Pro
1.6
independent
Sep 20, 2026
66Mercury 2
1.6
independent
Sep 20, 2026
69Qwen3 235B-A22Bthinking
1.3
mirror
Sep 8, 2026
70GPT-4.1 nano
1.2
mirror
Jul 23, 2026
71DeepSeek R1
1.1
independent
Sep 7, 2026
72GPT-4o mini
1.0
mirror
Jul 23, 2026
72nemotron-3-nano-30b-a3b
1.0
mirror
Sep 8, 2026
74o3-minihigh-effort
0.9
independent
Sep 7, 2026
74Qwen3 32B
0.9
mirror
Sep 8, 2026
74Qwen3 14B
0.9
mirror
Sep 8, 2026
77Qwen3 8B
0.8
mirror
Sep 8, 2026
78Llama-3.3-70B-Instruct
0.3
mirror
Jul 23, 2026

Bars on a 0–60 index scale. Scores link to model pages; every number keeps its source on the model page.