Skip to content

Benchmark

Nejumi 4: TOTAL_SCORE

Higher scores are better.

Models measured
114
Scores recorded
126
Sources
1
IQ weight
1.76

Score distribution

0.73
Mean
0.743
Median
0.097
Std dev
0.589
10th percentile
0.831
90th percentile
0.404
Min
0.4040.872
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.76 for chat models, measured over 114 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Opus 5anthropic/claude-opus-5: adaptive-thinking-max· Nejumi Leaderboard 4 data docs0.872
2Claude Fable 5anthropic/claude-fable-5: adaptive-thinking-max with fallback to Opus 4.8· Nejumi Leaderboard 4 data docs0.87
3Claude Opus 4.8anthropic/claude-opus-4.8: adaptive-thinking-xhigh· Nejumi Leaderboard 4 data docs0.852
4Claude Opus 4.7anthropic/claude-opus-4.7: adaptive-thinking-xhigh· Nejumi Leaderboard 4 data docs0.851
5GPT-5.6 Solopenai/gpt-5.6-sol: max-effort· Nejumi Leaderboard 4 data docs0.85
6GPT-5.6 Terraopenai/gpt-5.6-terra: max-effort· Nejumi Leaderboard 4 data docs0.844
7Kimi K3moonshotai/kimi-k3: reasoning-max· Nejumi Leaderboard 4 data docs0.843
8Gemini 3.1 Pro Previewgoogle/gemini-3.1-pro-preview· Nejumi Leaderboard 4 data docs0.843
9GPT 5.5openai/gpt-5.5-2026-04-23: xhigh-effort· Nejumi Leaderboard 4 data docs0.841
10GPT 5.4openai/gpt-5.4-2026-03-05: high-effort· Nejumi Leaderboard 4 data docs0.84
11Claude Opus 4.6anthropic/claude-opus-4-6: extended-thinking· Nejumi Leaderboard 4 data docs0.839
12GPT 5.5openai/gpt-5.5-2026-04-23: high-effort· Nejumi Leaderboard 4 data docs0.839
13Gemini 3.6 Flashgoogle/gemini-3.6-flash· Nejumi Leaderboard 4 data docs0.831
14Qwen3.6 Max Previewqwen/qwen3.6-max-preview: openrouter-reasoning· Nejumi Leaderboard 4 data docs0.83
15GPT 5.4openai/gpt-5.4-2026-03-05: xhigh-effort· Nejumi Leaderboard 4 data docs0.829
16GPT 5.2openai/gpt-5.2-2025-12-11: xhigh-effort· Nejumi Leaderboard 4 data docs0.829
17Gemini 3.5 Flashgoogle/gemini-3.5-flash· Nejumi Leaderboard 4 data docs0.825
18Claude Sonnet 4.6anthropic/claude-sonnet-4.6: extended-thinking· Nejumi Leaderboard 4 data docs0.823
19Qwen3.5 397B A17BQwen/Qwen3.5-397B-A17B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.819
20GLM 5.2zai-org/GLM-5.2: reasoning-max· Nejumi Leaderboard 4 data docs0.816
21Grok 4.5xai/grok-4.5: reasoning-high· Nejumi Leaderboard 4 data docs0.816
22Gemini 3 Flashgoogle/gemini-3-flash-preview· Nejumi Leaderboard 4 data docs0.816
23GPT-5.6 Lunaopenai/gpt-5.6-luna: max-effort· Nejumi Leaderboard 4 data docs0.815
24Gemini 3 Pro Previewgoogle/gemini-3-pro-preview· Nejumi Leaderboard 4 data docs0.813
25Qwen3.5 122B-A10BQwen/Qwen3.5-122B-A10B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.809
26GPT 5.1openai/gpt-5.1-2025-11-13: high-effort· Nejumi Leaderboard 4 data docs0.808
27Gemma 4 31B ITgoogle/gemma-4-31b-it· Nejumi Leaderboard 4 data docs0.808
28DeepSeek V4 Prodeepseek-ai/DeepSeek-V4-Pro: reasoning-max· Nejumi Leaderboard 4 data docs0.807
29Claude Opus 4.5anthropic/claude-opus-4.5-20251125: extended-thinking· Nejumi Leaderboard 4 data docs0.806
30MiniMax M3MiniMaxAI/MiniMax-M3: reasoning-enabled· Nejumi Leaderboard 4 data docs0.806
31Qwen3.5 27BQwen/Qwen3.5-27B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.805
32GLM 5.2z-ai/glm-5.2: openrouter-reasoning-xhigh· Nejumi Leaderboard 4 data docs0.804
33Claude Opus 4.1anthropic/claude-opus-4-1-20250805: extended-thinking· Nejumi Leaderboard 4 data docs0.799
34Qwen3.6 35B A3BQwen/Qwen3.6-35B-A3B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.798
35GPT 5openai/gpt-5-2025-08-07: high-effort· Nejumi Leaderboard 4 data docs0.797
36DeepSeek V4 Prodeepseek/deepseek-v4-pro: thinking-max· Nejumi Leaderboard 4 data docs0.796
37Qwen3.6 27BQwen/Qwen3.6-27B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.796
38Claude Sonnet 4.5anthropic/claude-sonnet-4-5-20250929: extended-thinking· Nejumi Leaderboard 4 data docs0.795
39Claude Sonnet 4anthropic/claude-sonnet-4-20250514: extended-thinking· Nejumi Leaderboard 4 data docs0.792
40DeepSeek Reasonerdeepseek/DeepSeek-V3.2 (Thinking Mode)· Nejumi Leaderboard 4 data docs0.79
41Qwen3.5 35B A3BQwen/Qwen3.5-35B-A3B: reasoning-enabled· Nejumi Leaderboard 4 data docs0.79
42DeepSeek V3.2deepseek-ai/DeepSeek-V3.2: reasoning-enabled· Nejumi Leaderboard 4 data docs0.789
43GLM 5zai-org/GLM-5: reasoning-enabled· Nejumi Leaderboard 4 data docs0.788
44Claude Haiku 4.5anthropic/claude-haiku-4-5-20251001: extended-thinking· Nejumi Leaderboard 4 data docs0.788
45OpenAI o3openai/o3-2025-04-16: high-effort· Nejumi Leaderboard 4 data docs0.788
46Gemma 4 26B A4Bgoogle/gemma-4-26B-A4B-it· Nejumi Leaderboard 4 data docs0.787
47Grok 4grok-4· Nejumi Leaderboard 4 data docs0.781
48Claude Opus 4anthropic/claude-opus-4-20250514: no-thinking· Nejumi Leaderboard 4 data docs0.78
49Kimi K2.5moonshotai/Kimi-K2.5: reasoning-enabled· Nejumi Leaderboard 4 data docs0.779
50Qwen 3 235B A22B Thinking 2507Qwen/Qwen3-235B-A22B-Thinking-2507: reasoning-enabled· Nejumi Leaderboard 4 data docs0.778

Showing 1–50 of 126 results.

Data credits: Nejumi Leaderboard 4, Weights & Biases Japan; public publisher-selected numeric evaluation results