NYT Connections Extended — leaderboard
Extended version: 940 puzzles with extra trick words added that don't fit any category. Models get one attempt. Significantly harder than the original.
Metric: Score (%). Source: github.com. Status: saturated. 92 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 97.8 |
| 2 | GPT-5.5 (xHigh) | 97.8 |
| 3 | GPT-5.5 (High) | 97.2 |
| 4 | GPT-5.4 (xHigh) | 96 |
| 5 | Gemini 3 Pro (Preview) | 95.2 |
| 6 | Gemini 3.5 Flash | 94.2 |
| 7 | GPT-5.4 (High) | 93 |
| 8 | GPT-5.5 (Medium) | 91.8 |
| 9 | Claude Opus 4.6 (Thinking, High) | 91.5 |
| 10 | Kimi K2.6 | 86.5 |
| 11 | Grok 4.20 0309 (Reasoning) | 85.5 |
| 12 | GPT-5.2 (xHigh) | 85 |
| 13 | Qwen 3.7 Max | 84.8 |
| 14 | GLM-5.1 | 84.2 |
| 15 | Grok 4.1 Fast (Reasoning) | 83.8 |
Interactive version: theaggregate.ai/benchmark?slug=nyt-connections-extended · How the rankings work · Data refreshed daily, snapshot 2026-07-22.