Artificial Analysis

Artificial Analysis ITBench-AA

An independently evaluated IT-operations benchmark from Artificial Analysis.

In index
Agentic
Best in agentic
Source ↗

Models24

Top result56.2%GPT-5.6 Sol

Top-3 spread3.7 pts

Year2026

Result and price

23 of 24 results

Results

24 results

#ModelResult

01GPT-5.6 Sol56.2%

02Step 5 Preview55.6%

03Gemini 3.8 Flash52.5%

04GLM-5.3-Flash51.2%

05GPT-5.6 Terra51.0%

06Claude Fable 5.149.5%

07GPT-6 Sol49.4%

08GPT-6 Astra48.6%

09Kimi K347.7%

10DeepSeek V4.1 Flash 101746.9%

11Claude Opus 4.7 (Adaptive)46.7%

12GLM-5.346.1%

13GPT-5.545.8%

14GLM-5.242.7%

15Qwen3.7 Max42.5%

16Grok 4.742.1%

17GPT-5.6 Luna40.3%

17Gemini 3.5 Flash40.3%

17GLM-5.140.3%

20Claude Opus 5.538.2%

20MiMo-V2.5-Pro38.2%

22Gemma 4 31B37.3%

23Qwen3.5-27B35.5%

24GPT-5.4 mini35.2%