Skip to content

Benchmark

NYT Connections – Newest 100 puzzles, extended version (Lech Mazur)

Scores are percentages.

Models measured
106
Scores recorded
138
Sources
1
IQ weight
2.05

Score distribution

61.5%
Mean
70.9%
Median
30%
Std dev
14.6%
10th percentile
94.5%
90th percentile
5.5%
Min
5.5%99.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.05 for chat models, measured over 106 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Qwen3.7 Max· Lech Mazur suite data78.9%
52GPT 5.2xhigh reasoning· Lech Mazur suite data78.6%
53Claude Opus 4.616k· Lech Mazur suite data78.4%
54GLM 5.1· Lech Mazur suite data77.6%
55Grok 4.20Reasoning· Lech Mazur suite data77.4%
56Grok 4.5high reasoning· Lech Mazur suite data77.2%
57Muse Spark 1.2high reasoning· Lech Mazur suite data77.2%
58Grok 4.6xhigh reasoning· Lech Mazur suite data77.1%
59Step 5 Previewhigh reasoning· Lech Mazur suite data75.4%
60Grok 4.1 Fastreasoning· Lech Mazur suite data75.2%
61Grok 4.6high reasoning· Lech Mazur suite data74.6%
62Muse Spark Openhigh reasoning· Lech Mazur suite data74.3%
63Claude Sonnet 4.6high reasoning· Lech Mazur suite data74.1%
64GLM 5.2high reasoning· Lech Mazur suite data72.8%
65Grok 4.7high reasoning· Lech Mazur suite data72.8%
66GPT-5.6 Terrahigh reasoning· Lech Mazur suite data72.4%
67GLM 5.3high reasoning· Lech Mazur suite data72.2%
68GPT-5.2 Pro· Lech Mazur suite data71.8%
69Qwen3.7 Plus· Lech Mazur suite data70.9%
70Claude Sonnet 5high reasoning· Lech Mazur suite data69.9%
71Claude Sonnet 4.632k· Lech Mazur suite data69.4%
72Qwen3.6 Max Preview· Lech Mazur suite data69.4%
73GLM 5· Lech Mazur suite data67.5%
74GPT 5.2high reasoning· Lech Mazur suite data66.2%
75GPT-5.6 Lunahigh reasoning· Lech Mazur suite data66.1%
76Gemma 4 31B ITreasoning· Lech Mazur suite data65.9%
77GPT-6 Lunahigh reasoning· Lech Mazur suite data65.8%
78Claude Haiku 5.5high reasoning· Lech Mazur suite data64.9%
79Hy4 preview· Lech Mazur suite data64.2%
80Kimi K2.5thinking· Lech Mazur suite data63.7%
81MiniMax M3· Lech Mazur suite data63.4%
82GPT-5.4 minixhigh reasoning· Lech Mazur suite data62.5%
83Gemini 3 Flash· Lech Mazur suite data61.8%
84Gemini 3.5 Flash Litehigh reasoning· Lech Mazur suite data60.7%
85DeepSeek-V4-Pro Preview· Lech Mazur suite data59.9%
86Qwen3.6 Plus· Lech Mazur suite data57.1%
87GPT 5.2medium reasoning· Lech Mazur suite data55.8%
88Qwen 3.8 27B· Lech Mazur suite data52%
89Claude Opus 4.5thinking, 16k· Lech Mazur suite data51.6%
90Qwen3.5 397B A17B· Lech Mazur suite data47.1%
91GPT 5.2low reasoning· Lech Mazur suite data46.4%
92Grok 4.3· Lech Mazur suite data45.6%
93Qwen3.5 122B-A10B· Lech Mazur suite data45.4%
94Claude Sonnet 4.616k· Lech Mazur suite data45.3%
95Qwen3.5 27B· Lech Mazur suite data44.6%
96Claude Sonnet 4.5no reasoning· Lech Mazur suite data43.6%
97Claude Sonnet 4.5thinking, 16k· Lech Mazur suite data43%
98Claude Opus 4.5no reasoning· Lech Mazur suite data42%
99Hy3 (Deep Infra)high· Lech Mazur suite data40.6%
100Claude Sonnet 4.6no reasoning· Lech Mazur suite data40.4%

Showing 51–100 of 138 results.