Gakucho Benchmark - Mathematics Open-Ended: leaderboard
Metric: Character-level F1 (0-100, the paper's 0-1 score times 100) of the answers to the 10 open-ended Mathematics items against the official correct answers, of the officially released Grade 9 items of Japan's 2022 National Assessment of Academic Ability (original exam layouts and diagrams, Japanese prompts, nationwide student response statistics), zero-shot with the item images; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 69.5 |
| 2 | Claude 3.7 Sonnet (20250219) | 66.3 |
| 3 | GPT-5 | 62.3 |
| 4 | Gemini 2.5 Pro | 55.8 |
| 5 | Claude Sonnet 4 (20250514) | 54.8 |
| 6 | Gemini 2.5 Flash | 54.2 |
| 7 | GPT-4o | 49.2 |
| 8 | GPT-4o Mini | 46.1 |
| 9 | Gemini 3.1 Pro (Preview) | 35.8 |
| 10 | Gemma 3 12B (IT) | 24.6 |
| 11 | Qwen 2.5 VL 7B Instruct | 19.5 |
Interactive version: theaggregate.ai/benchmark?slug=gakucho-benchmark-mathematics-open-ended · How It Works · Data refreshed daily, snapshot 2026-10-07.