Skip to content

Benchmark

NYT Connections – Extended Version (Lech Mazur)

Scores are percentages.

Models measured
106
Scores recorded
138
Sources
1
IQ weight
2.44

Score distribution

63.5%
Mean
75.1%
Median
30.5%
Std dev
14.5%
10th percentile
94.4%
90th percentile
4.1%
Min
4.1%99.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.44 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
51Muse Spark 1.1high reasoning· Lech Mazur suite data84.9%
52GPT 5.2xhigh reasoning· Lech Mazur suite data83.6%
53Hy 4high· Lech Mazur suite data83.5%
54Gemini 3 Flash· Lech Mazur suite data83.1%
55Claude Sonnet 4.6high reasoning· Lech Mazur suite data80.9%
56Claude Sonnet 5.5high reasoning· Lech Mazur suite data80.5%
57Grok 4.6xhigh reasoning· Lech Mazur suite data80%
58Grok 4.5high reasoning· Lech Mazur suite data79.9%
59GPT-5.2 Pro· Lech Mazur suite data79.3%
60Muse Spark 1.2high reasoning· Lech Mazur suite data79.2%
61Grok 4.6high reasoning· Lech Mazur suite data79%
62GPT-5.6 Terrahigh reasoning· Lech Mazur suite data78.4%
63Muse Spark Openhigh reasoning· Lech Mazur suite data78.4%
64GLM 5.1· Lech Mazur suite data77.7%
65Grok 4.7high reasoning· Lech Mazur suite data76.8%
66Step 5 Previewhigh reasoning· Lech Mazur suite data76.6%
67Claude Opus 4.616k· Lech Mazur suite data76.4%
68Claude Sonnet 4.632k· Lech Mazur suite data76.4%
69Claude Sonnet 5high reasoning· Lech Mazur suite data75.1%
70GLM 5· Lech Mazur suite data74.8%
71Qwen3.7 Plus· Lech Mazur suite data74.8%
72GLM 5.2high reasoning· Lech Mazur suite data74.3%
73GLM 5.3high reasoning· Lech Mazur suite data74.2%
74Qwen3.6 Max Preview· Lech Mazur suite data74.1%
75Gemma 4 31B ITreasoning· Lech Mazur suite data70.6%
76Kimi K2.5thinking· Lech Mazur suite data69.9%
77GPT-5.6 Lunahigh reasoning· Lech Mazur suite data69.4%
78GPT-6 Lunahigh reasoning· Lech Mazur suite data68.7%
79Hy4 preview· Lech Mazur suite data68.2%
80GPT 5.2high reasoning· Lech Mazur suite data68.1%
81DeepSeek-V4-Pro Preview· Lech Mazur suite data67.3%
82Claude Haiku 5.5high reasoning· Lech Mazur suite data65.7%
83MiniMax M3· Lech Mazur suite data65.1%
84GPT-5.4 minixhigh reasoning· Lech Mazur suite data61.8%
85GPT 5.2medium reasoning· Lech Mazur suite data60.6%
86Gemini 3.5 Flash Litehigh reasoning· Lech Mazur suite data60.4%
87Qwen3.6 Plus· Lech Mazur suite data60.3%
88Qwen3.5 397B A17B· Lech Mazur suite data58.9%
89Grok 4.3· Lech Mazur suite data55.2%
90Qwen 3.8 27B· Lech Mazur suite data54.5%
91GPT 5.2low reasoning· Lech Mazur suite data54.2%
92Claude Opus 4.5thinking, 16k· Lech Mazur suite data52.5%
93Qwen3.5 122B-A10B· Lech Mazur suite data51.7%
94Claude Opus 4.5no reasoning· Lech Mazur suite data49.4%
95Claude Sonnet 4.616k· Lech Mazur suite data48%
96Qwen3.5 27B· Lech Mazur suite data47.9%
97Claude Sonnet 4.6no reasoning· Lech Mazur suite data44.8%
98Qwen3.7 Flash· Lech Mazur suite data43.8%
99Claude Opus 4.6no reasoning· Lech Mazur suite data43.2%
100Qwen3.6 35B A3B· Lech Mazur suite data41.6%

Showing 51–100 of 138 results.