Kangaroo 2025 3-4
Math · 2025-10-20
This MathArena track evaluates models on the Kangaroo 2025 3-4 problem set. Scores report the percentage of problems answered correctly.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5.5 | 89.6 |
| GPT-5.4 | 83.3 |
| Kimi K3 | 77.1 |
| Gemini 3.1 Pro Preview | 76.0 |
| GPT-5.2 | 74.0 |
| Gemini 3.5 Flash | 72.9 |
| Gemini 3.6 Flash | 68.8 |
| Qwen3.5 397B A17B | 68.8 |
| Gemini 3 Flash Preview | 66.7 |
| Gemini 3 Pro Preview | 66.7 |
| GPT-5 Mini | 66.7 |
| GPT-5.1 | 65.6 |
| Kimi K2.5 | 65.6 |
| Gemini 2.5 Pro | 64.6 |
| Sonnet 4.5 | 62.5 |
| Opus 4.8 | 60.4 |
| GPT-5 | 60.4 |
| Qwen3-VL-235B-A22B-Instruct | 58.3 |
| Grok 4 | 52.1 |
| Opus 4.6 | 50.0 |
| GLM-4.5V | 45.8 |
| Grok 4.1 Fast | 39.6 |
| Grok 4 Fast | 32.3 |