Leaderboard bars
20 ranked models · higher is better
View accessible chart data
| Model | Rank | Provider | Score |
|---|---|---|---|
| GPT-5.6 Sol | #1 | OpenAI | 92.2% |
| Kimi K3 | #2 | Moonshot AI | 91.2% |
| GPT-5.5 Pro | #3 | OpenAI | 90.1% |
| Claude Mythos 5 | #4 | Anthropic | 88% |
| GPT-5.6 Terra | #5 | OpenAI | 87.5% |
| Claude Sonnet 5 | #6 | Anthropic | 84.7% |
| GPT-5.5 | #7 | OpenAI | 84.4% |
| Claude Opus 4.8 | #8 | Anthropic | 84.3% |
| Claude Opus 4.6 | #9 | Anthropic | 83.7% |
| MiniMax M3 | #10 | MiniMax | 83.5% |
| DeepSeek V4 Pro | #11 | DeepSeek | 83.4% |
| GPT-5.6 Luna | #12 | OpenAI | 83.3% |
| Kimi K2.6 | #13 | Moonshot AI | 83.2% |
| GPT-5.4 | #14 | OpenAI | 82.7% |
| Claude Opus 4.7 | #15 | Anthropic | 79.3% |
| GLM-5.1 | #16 | Z.AI | 68% |
| GPT-5.2 | #17 | OpenAI | 65.8% |
| Qwen3.5 122B | #18 | Alibaba Cloud | 63.8% |
| Qwen3.5 397B | #19 | Alibaba Cloud | 62% |
| Qwen3.5 27B | #20 | Alibaba Cloud | 61% |