Artificial Analysis Intelligence Index v5.0
Indexes · 2026-10-27
Artificial Analysis Intelligence Index v5.0, launched October 27, 2026, weights nine evaluations in four equal areas: knowledge work (AA-Briefcase v1.1 15%, AutomationBench-AA 5%, GDP.pdf 5%), coding (FrontierCode 2.0 15%, Terminal-Bench 4.0 10%), science (Terminal-Bench-Science 0.1 15%, the 40-question CritPt v2 hard subset 10%) and general knowledge (AA-Omniscience v1.1 15%, split 10% accuracy and 5% non-hallucination, and HLE-Diamond 10%). It drops GDPval-AA v2.1, SciCode and AA-LCR v1.1 and changes three other components, so its scores are a separate scale from v4.3 and v4.3.2.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-6 Astra | 56.9 |
| GPT-6.1 Sol | 54.5 |
| Claude Sonnet 5.5 | 54.2 |
| GPT-6.1 Terra | 51.0 |
| Gemini 4 Argon | 50.4 |
| Muse Wonder Full | 49.7 |
| Kimi K3.1 Pro | 46.0 |
| Muse Wonder | 43.3 |
| Muse Spark 1.3 | 36.4 |
| Claude Haiku 5.5 | 36.2 |
| Grok 4.7 | 35.0 |
| Qwen3.8 Max (0902) | 33.6 |
| DeepSeek V4.1 Flash 1017 | 32.5 |
| GLM-5.3 | 31.7 |
| Gemini 3.8 Flash | 30.5 |
| MiMo-V2.6-Pro | 29.9 |
| Ling 3.1 Flash 1019 | 29.5 |
| MiniMax M3.1 Flash | 28.9 |
| Step 5 Preview | 28.5 |
| GPT-6 Luna | 27.8 |
| MiMo-V2.6-Flash | 26.7 |
| Mistral Large 4 Preview | 25.3 |
| Muse Spark Open | 24.7 |
| DeepSeek V4 Pro 0813 | 23.1 |
| Qwen3.8 27B | 20.1 |
| MiniMax M3 | 18.7 |
| K2 Horizon 375B A23B | 17.8 |
| Beam 1.0 501B A23B | 17.7 |
| Nemotron 3.5 Super VL 120B A12B (precision unspecified) | 13.6 |
| Inkling-Small | 12.9 |
| Inkling | 12.5 |
| Nemotron 3 Ultra 550B A55B | 11.8 |
| Gemini 3.5 Flash-Lite | 11.7 |
| Muse Glimmer 30B | 6.3 |
| Command A+ (May 2026, BF16) | 6.2 |
| Nemotron 3 Nano 4B | 1.8 |