SAGE - Partial Differential Equations
Math · 2025-10-08
The Partial Differential Equations split of SAGE. This child benchmark reports source-level scores for the PDEs mathematical problem category.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemma 4 31B IT | 50.8 |
| Gemini 3.5 Flash-Lite | 49.2 |
| Opus 4.5 | 47.6 |
| Opus 4.7 | 47.6 |
| Gemini 3.1 Flash-Lite Preview | 47.6 |
| Gemini 3 Pro Preview | 47.6 |
| GPT-5.6 Sol | 47.6 |
| MiniMax M3 | 47.6 |
| GPT-5.2 | 46.0 |
| Kimi K3 | 46.0 |
| Claude Fable 5 | 44.4 |
| Opus 4.8 | 44.4 |
| Kimi K2.5 | 44.4 |
| Sonnet 5 | 42.9 |
| Gemini 3.5 Flash | 42.9 |
| GPT-5.1 | 42.9 |
| Opus 4.6 | 41.3 |
| Sonnet 4.6 | 41.3 |
| Gemini 3.6 Flash | 41.3 |
| Gemini 3 Flash Preview | 41.3 |
| MiMo-V2.5 | 41.3 |
| Mistral Medium 3.5 | 41.3 |
| Qwen3.6 27B | 41.3 |
| Claude Opus 5 | 39.7 |
| Gemini 2.5 Flash Preview (09-2025) | 39.7 |
| Gemini 3.1 Pro Preview | 39.7 |
| GPT-5.4 Mini | 39.7 |
| GPT-5.4 | 39.7 |
| GPT-5.5 | 39.7 |
| GPT-5 | 39.7 |
| Kimi K2.6 | 39.7 |
| Muse Spark 1.1 | 39.7 |
| Qwen3.5-Flash | 39.7 |
| GPT-5.6 Luna | 38.1 |
| GPT-5.6 Terra | 38.1 |
| Qwen3.6 Plus (2026-04-02) | 38.1 |
| GPT-5 Mini | 36.5 |
| Sonnet 4 | 34.9 |
| Grok 4.20 | 34.9 |
| Llama 4 Maverick Instruct | 34.9 |