llm.ing

AA Coding Index

run by Artificial Analysisunit: indexDisplay with attribution; no redistribution73 modelssource ↗
AA Intelligence IndexAA Coding IndexAA Agentic IndexMedian output speedMedian time to first token
#ModelScoreObserved
1GPT-5.6 Sol
77.4
mirror
Jul 23, 2026
2GPT-5.6 Terra
76.7
mirror
Jul 23, 2026
3Claude Fable 5
76.5
mirror
Jul 23, 2026
4Kimi K3
76.2
mirror
Jul 23, 2026
5GPT-5.5
74.9
independent
Jul 24, 2026
6Claude Opus 4.8
74.3
mirror
Jul 23, 2026
7Claude Opus 4.7
73.6
mirror
Jul 23, 2026
8Grok 4.5
72.4
mirror
Jul 23, 2026
9Claude Sonnet 5
71.5
independent
Jul 24, 2026
10GPT-5.6 Luna
71.4
independent
Jul 24, 2026
11Muse Spark 1.1
71.3
mirror
Jul 23, 2026
12GPT-5.4
71.1
mirror
Jul 23, 2026
13Gemini 3.5 Flash
70.1
independent
Jul 24, 2026
14Gemini 3.6 Flash
69.2
mirror
Jul 23, 2026
15Gemini 3.1 Pro Preview
68.8
mirror
Jul 23, 2026
15GLM-5.2
68.8
mirror
Jul 23, 2026
17Qwen3.7 Max
66.0
mirror
Jul 23, 2026
18Claude Sonnet 4.6
63.0
mirror
Jul 23, 2026
19Kimi K2.6
61.8
mirror
Jul 23, 2026
20Kimi K2.7 Code
60.8
mirror
Jul 23, 2026
21DeepSeek V4 Pro
59.4
mirror
Jul 23, 2026
22MiniMax-M3
58.6
independent
Jul 24, 2026
23DeepSeek V4 Flash
56.2
mirror
Jul 23, 2026
24GPT-5.4 nano
56.1
mirror
Jul 23, 2026
24GPT-5.4 mini
56.1
mirror
Jul 23, 2026
26Qwen3.7 Plus
55.9
mirror
Jul 23, 2026
27GLM-5.1
55.8
mirror
Jul 23, 2026
28Qwen3.6 Plus
54.5
mirror
Jul 23, 2026
29Qwen3.6 27B
53.7
mirror
Jul 23, 2026
30MiniMax-M2.7
52.6
independent
Jul 24, 2026
31Claude Sonnet 4.5 (latest)
52.1
mirror
Jul 23, 2026
32Grok Build 0.1
51.5
independent
Jul 24, 2026
33GPT-5.1
49.4
independent
Jul 24, 2026
34Gemini 3.5 Flash Lite
49.3
independent
Jul 24, 2026
34Nemotron 3 Ultra 550B A55B
49.3
mirror
Jul 23, 2026
36Qwen3.5 397B-A17B
48.2
independent
Jul 24, 2026
37Kimi K2.5
46.8
mirror
Jul 23, 2026
38GLM-4.6
45.8
mirror
Jul 23, 2026
39Qwen3.5 122B-A10B
45.7
independent
Jul 24, 2026
40GLM-4.7
45.3
mirror
Jul 23, 2026
41DeepSeek V3.2
44.2
mirror
Jul 23, 2026
42Claude Haiku 4.5 (latest)
43.9
mirror
Jul 23, 2026
43Gemma 4 31B IT
43.4
mirror
Jul 23, 2026
44Grok 4.3
42.2
mirror
Jul 23, 2026
45Qwen3.6 35B-A3B
41.9
mirror
Jul 23, 2026
46o1
39.7
mirror
Jul 23, 2026
47Step 3.7 Flash
39.6
mirror
Jul 23, 2026
48Gemma 4 26B A4B IT
39.3
mirror
Jul 23, 2026
49GPT-5
37.8
independent
Jul 24, 2026
50Nemotron 3 Super
37.7
mirror
Jul 23, 2026
51Qwen3.5 35B-A3B
37.0
mirror
Jul 23, 2026
52Gemini 3.1 Flash Lite Preview
34.7
mirror
Jul 23, 2026
53Gemini 2.5 Pro
33.3
independent
Jul 24, 2026
54Devstral 2
31.3
mirror
Jul 23, 2026
55Mercury 2
31.1
independent
Jul 24, 2026
56DeepSeek R1
24.6
independent
Jul 24, 2026
57GPT-4o
24.2
mirror
Jul 23, 2026
58Qwen3 235B-A22Bthinking
22.1
mirror
Jul 23, 2026
59GPT-4 Turbo
21.5
mirror
Jul 23, 2026
60Kimi K2 Thinkingthinking
21.0
mirror
Jul 23, 2026
61GPT-4.1 mini
20.2
mirror
Jul 23, 2026
62Qwen3-Next 80B-A3B (Thinking)thinking
17.4
mirror
Jul 23, 2026
63o3-minihigh-effort
16.3
independent
Jul 24, 2026
64GPT-5 Mini
15.6
independent
Jul 24, 2026
65Qwen3 32B
15.3
mirror
Jul 23, 2026
66nemotron-3-nano-30b-a3b
14.4
mirror
Jul 23, 2026
67Qwen3 14B
13.8
mirror
Jul 23, 2026
68GPT-4
13.1
mirror
Jul 23, 2026
69Llama-3.3-70B-Instruct
11.9
mirror
Jul 23, 2026
70GPT-4o mini
11.4
mirror
Jul 23, 2026
71GPT-4.1 nano
11.1
mirror
Jul 23, 2026
72GPT-3.5-turbo
10.7
mirror
Jul 23, 2026
73Qwen3 8B
9.0
mirror
Jul 23, 2026

Bars on a 0–77.4 index scale. Scores link to model pages; every number keeps its source on the model page.