Pencil Puzzle Bench - Country — leaderboard
Metric: Direct-ask Success Rate (%). Source: ppbench.com. 51 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 6.7 |
| 2 | Gemini 2.5 Pro | 0 |
| 3 | GPT-4o | 0 |
| 4 | Claude Sonnet 4.6 | 0 |
| 5 | GPT-4.1 | 0 |
| 6 | GPT-5.2 | 0 |
| 7 | Claude Sonnet 4.5 | 0 |
| 8 | GPT-OSS-120B | 0 |
| 9 | Claude Opus 4.6 | 0 |
| 10 | Claude Sonnet 4.5 (Thinking) | 0 |
| 11 | GPT-3.5 Turbo | 0 |
| 12 | Grok 4.1 Fast (Reasoning) | 0 |
| 13 | DeepSeek V3.2 | 0 |
| 14 | Kimi K2.5 | 0 |
| 15 | Qwen 3 Next 80B A3B (Thinking) | 0 |
Interactive version: theaggregate.ai/benchmark?slug=pencil-puzzle-bench-country · How the rankings work · Data refreshed daily, snapshot 2026-07-22.