Physics on Physics tasks
52.5ScoreClaude-Sonnet-4.5-Think
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Claude-Sonnet-4.5-ThinkGeneration Mode=LLM-in-Sandbox2026.01 | 52.5 | 5.8 | |
| Claude-Sonnet-4.5-ThinkGeneration Mode=Standard LLM2026.01 | 46.7 | — | |
| GPT-5Generation Mode=LLM-in-Sandbox2026.01 | 45.8 | 4.8 | |
| DeepSeek-V3.2-ThinkingGeneration Mode=LLM-in-Sandbox2026.01 | 45.2 | 7.6 | |
| Kimi-K2-ThinkingGeneration Mode=LLM-in-Sandbox2026.01 | 42.3 | 1.7 | |
| GPT-5Generation Mode=Standard LLM2026.01 | 41 | — | |
| Kimi-K2-ThinkingGeneration Mode=Standard LLM2026.01 | 40.6 | — | |
| DeepSeek-V3.2-ThinkingGeneration Mode=Standard LLM2026.01 | 37.6 | — | |
| MiniMax-M2Generation Mode=LLM-in-Sandbox2026.01 | 35.8 | 5.5 | |
| MiniMax-M2Generation Mode=Standard LLM2026.01 | 30.3 | — | |
| Qwen3-Coder-30B-A3BGeneration Mode=LLM-in-Sandbox2026.01 | 28.9 | 5 | |
| Qwen3-4B-Instruct-2507Generation Mode=LLM-in-Sandbox2026.01 | 25.6 | 2.7 | |
| Qwen3-Coder-30B-A3BGeneration Mode=Standard LLM2026.01 | 23.9 | — | |
| Qwen3-4B-Instruct-2507Generation Mode=Standard LLM2026.01 | 22.9 | — |