Artificial Analysis ITBench-AA leaderboard
14 ranked models · higher is better
View accessible chart data
| Model | Rank | Provider | Score |
|---|---|---|---|
| GPT-5.6 Sol | #1 | OpenAI | 56.2% |
| GPT-5.6 Terra | #2 | OpenAI | 51% |
| Kimi K3 | #3 | Moonshot AI | 47.7% |
| Claude Opus 4.7 (Adaptive) | #4 | Anthropic | 46.7% |
| GPT-5.5 | #5 | OpenAI | 45.8% |
| GLM-5.2 | #6 | Z.AI | 42.7% |
| Qwen3.7-Max | #7 | Alibaba Cloud | 42.5% |
| GPT-5.6 Luna | #8 | OpenAI | 40.3% |
| DeepSeek V4 Pro (Max) | #9 | DeepSeek | 38.3% |
| MiMo-V2.5-Pro | #10 | Xiaomi | 38.2% |
| Gemma 4 31B | #11 | 37.3% | |
| DeepSeek V4 Flash (Max) | #12 | DeepSeek | 31.5% |
| Gemini 3.1 Pro | #13 | 30.3% | |
| GPT-OSS 120B | #14 | OpenAI | 5.6% |