GR-Ben (Thinking): leaderboard
Metric: Average F1 (%) over the nine GR-Ben subsets (four scientific: biology, physics, chemistry and computer science from MMLU-Pro; five logical: abductive, analogical, mixed-form, deductive and inductive), F1 being the harmonic mean of the accuracy on flawed solutions (earliest erroneous step named) and on fully correct solutions; LLMs prompted as critics with slow thinking enabled (GPT-5.2 at xhigh and Gemini 3 Flash at high reasoning effort, 32,768 output tokens); higher is better. Source: arxiv.org. Saturation forecast: Around September 2028. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (High) | 61.1 |
| 2 | DeepSeek V3.2 (Thinking) | 56.7 |
| 3 | GPT-5.2 (xHigh) | 54.9 |
| 4 | Qwen 3 32B | 45.4 |
| 5 | Qwen 3 14B | 42.3 |
| 6 | Qwen 3 8B | 39.8 |
| 7 | Qwen 3 4B | 32.5 |
Interactive version: theaggregate.ai/benchmark?slug=gr-ben-thinking · How It Works · Data refreshed daily, snapshot 2026-10-07.