Skip to content

Benchmark

Horangi 4 subset: ko_arc_agi

Higher scores are better.

Models measured
79
Scores recorded
83
Sources
1
IQ weight
1.01

Score distribution

0.419
Mean
0.438
Median
0.32
Std dev
0.01
10th percentile
0.906
90th percentile
0
Min
00.97
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.01 for chat models, measured over 79 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1GPT-5.6 Solgpt-5.6-sol: max-effort· Horangi Korean Leaderboard 4 data docs0.97
2GPT-5.6 Terragpt-5.6-terra: max-effort· Horangi Korean Leaderboard 4 data docs0.97
3Claude Fable 5claude-fable-5: high-effort· Horangi Korean Leaderboard 4 data docs0.96
4GPT 5.4gpt-5.4-2026-03-05: xhigh-effort· Horangi Korean Leaderboard 4 data docs0.95
5GPT 5.5gpt-5.5: xhigh-effort· Horangi Korean Leaderboard 4 data docs0.93
6GPT 5.2gpt-5.2-2025-12-11: xhigh-effort· Horangi Korean Leaderboard 4 data docs0.927
7Gemini 3.1 Pro Previewgemini-3.1-pro-preview: high-effort· Horangi Korean Leaderboard 4 data docs0.92
8Grok 4.20grok-4.20-beta· Horangi Korean Leaderboard 4 data docs0.907
9Grok 4.20grok-4.20· Horangi Korean Leaderboard 4 data docs0.906
10Gemini 3.5 Flashgemini-3.5-flash· Horangi Korean Leaderboard 4 data docs0.899
11Claude Sonnet 5claude-sonnet-5: high-effort· Horangi Korean Leaderboard 4 data docs0.85
12Grok 4.5grok-4.5· Horangi Korean Leaderboard 4 data docs0.84
13Claude Opus 4.6claude-opus-4-6: high-effort· Horangi Korean Leaderboard 4 data docs0.84
14GPT-5.6 Lunagpt-5.6-luna: max-effort· Horangi Korean Leaderboard 4 data docs0.82
15Qwen3.7 Plusqwen3.7-plus· Horangi Korean Leaderboard 4 data docs0.805
16Gemini 3 Pro Previewgemini-3-pro-preview: high-effort· Horangi Korean Leaderboard 4 data docs0.78
17GPT 5.1gpt-5.1-2025-11-13· Horangi Korean Leaderboard 4 data docs0.77
18Qwen3.7 Maxqwen3.7-max· Horangi Korean Leaderboard 4 data docs0.745
19Qwen3.6 Plus Previewqwen3.6-plus-preview· Horangi Korean Leaderboard 4 data docs0.707
20GPT-5.4 minigpt-5.4-mini-2026-03-17: xhigh-effort· Horangi Korean Leaderboard 4 data docs0.687
21Claude Sonnet 4.6claude-sonnet-4-6: high-effort· Horangi Korean Leaderboard 4 data docs0.68
22GPT 5gpt-5-2025-08-07· Horangi Korean Leaderboard 4 data docs0.68
23Qwen3.5 122B-A10Bqwen3.5-122b-a10b· Horangi Korean Leaderboard 4 data docs0.636
24Qwen3.5 27Bqwen3.5-27b· Horangi Korean Leaderboard 4 data docs0.623
25Qwen3.5 397B A17Bqwen3.5-397b-a17b· Horangi Korean Leaderboard 4 data docs0.615
26Claude Opus 4.5claude-opus-4-5-20251101: high-effort· Horangi Korean Leaderboard 4 data docs0.61
27Gemma 4 31B ITGemma-4-31B-IT-NVFP4: enable-thinking· Horangi Korean Leaderboard 4 data docs0.6
28Gemma 4 31B ITgemma-4-31B-it: enable-thinking· Horangi Korean Leaderboard 4 data docs0.58
29DeepSeek Reasonerdeepseek-v3.2: high-effort· Horangi Korean Leaderboard 4 data docs0.577
30GPT 5.4 Nanogpt-5.4-nano-2026-03-17: xhigh-effort· Horangi Korean Leaderboard 4 data docs0.565
31Grok 4.1 Fastgrok-4-1-fast-reasoning· Horangi Korean Leaderboard 4 data docs0.557
32GPT-5 minigpt-5-mini-2025-08-07· Horangi Korean Leaderboard 4 data docs0.53
33Qwen3.6 Flashqwen3.6-flash· Horangi Korean Leaderboard 4 data docs0.526
34GLM 5.2glm-5.2: high-effort· Horangi Korean Leaderboard 4 data docs0.5
35Kimi K2.5kimi-k2.5· Horangi Korean Leaderboard 4 data docs0.48
36MiniMax M2.7minimax-m2.7· Horangi Korean Leaderboard 4 data docs0.467
37MiniMax M2.5minimax-m2.5· Horangi Korean Leaderboard 4 data docs0.462
38Claude Opus 4.1claude-opus-4-1-20250805· Horangi Korean Leaderboard 4 data docs0.46
39GLM 5 Turboglm-5-turbo· Horangi Korean Leaderboard 4 data docs0.455
40Grok 4 Fastgrok-4-fast-reasoning· Horangi Korean Leaderboard 4 data docs0.45
41Qwen3.5 35B A3Bqwen3.5-35b-a3b· Horangi Korean Leaderboard 4 data docs0.439
42GLM 5glm-5· Horangi Korean Leaderboard 4 data docs0.438
43Claude Sonnet 4.5claude-sonnet-4-5-20250929: high-effort· Horangi Korean Leaderboard 4 data docs0.42
44MiMo V2 Promimo-v2-pro· Horangi Korean Leaderboard 4 data docs0.402
45Qwen3.5 9Bqwen3.5-9b· Horangi Korean Leaderboard 4 data docs0.394
46Claude Opus 4claude-opus-4-20250514: high-effort· Horangi Korean Leaderboard 4 data docs0.35
47Qwen 3 235B A22B Thinking 2507qwen3-235b-a22b-thinking-2507· Horangi Korean Leaderboard 4 data docs0.347
48MiMo V2 Flashmimo-v2-flash· Horangi Korean Leaderboard 4 data docs0.338
49GPT OSS 120Bgpt-oss-120b: high-effort· Horangi Korean Leaderboard 4 data docs0.3
50GLM 4.7glm-4.7· Horangi Korean Leaderboard 4 data docs0.299

Showing 1–50 of 83 results.

Data credits: Horangi 4, Weights & Biases Korea; public publisher-selected numeric evaluation results