Noometry

Reasoning benchmark

Mystery Game Puzzles leaderboard

As of October 2026, GPT-6 Astra has the highest published Mystery Game Puzzles score on Noometry at 84%, out of 74 models with results.

Last verified

About Mystery Game Puzzles

A description with primary sources is being prepared for this benchmark.

Category
Reasoning
Introduced
2026
Unit
Percent (random guessing ≈ 9.2%)
Official site
epoch.ai

Top 15 models

  1. GPT-6 Astra 84%
  2. GPT-6.1 Sol 80%
  3. Claude Opus 5.5 71%
  4. Claude Sonnet 5.5 65%
  5. Claude Opus 5 59%
  6. Claude Fable 5.1 58%
  7. GPT-5.6 Sol 58%
  8. GPT-5.5 56%
  9. GPT-6 Sol 56%
  10. Claude Fable 5 52%
  11. Gemini 3.8 Flash 47%
  12. DeepSeek V4.1 Flash 43%
  13. DeepSeek V4 Pro 43%
  14. Qwen3.8 Max 38%
  15. Gemini 3.7 Flash 37%

Sponsored placements are available on pages like this one. Advertise on Noometry

All results

#ModelProviderScoreSettingSourceDate
1GPT-6 Astra OpenAI84%maxEpoch AI2026-08-30
2GPT-6.1 Sol OpenAI80%maxEpoch AI2026-09-29
3Claude Opus 5.5 Anthropic71%maxEpoch AI2026-09-29
4Claude Sonnet 5.5 Anthropic65%maxEpoch AI2026-09-29
5Claude Opus 5 Anthropic59%maxEpoch AI2026-07-25
6Claude Fable 5.1 Anthropic58%maxEpoch AI2026-09-01
7GPT-5.6 Sol OpenAI58%maxEpoch AI2026-07-28
8GPT-5.5 OpenAI56%xhighEpoch AI2026-07-24
9GPT-6 Sol OpenAI56%maxEpoch AI2026-09-22
10Claude Fable 5 Anthropic52%maxEpoch AI2026-07-31
11Gemini 3.8 Flash Google47%highEpoch AI2026-09-02
12DeepSeek V4.1 Flash DeepSeek43%maxEpoch AI2026-10-08
13DeepSeek V4 Pro DeepSeek43%maxEpoch AI2026-08-19
14Qwen3.8 Max Alibaba (Qwen)38%xhighEpoch AI2026-08-05
15Gemini 3.7 Flash Google37%highEpoch AI2026-08-14
16GPT-5.4 OpenAI37%xhighEpoch AI2026-07-24
17Claude Opus 4.8 Anthropic36%maxEpoch AI2026-07-25
18Claude Sonnet 5 Anthropic35%maxEpoch AI2026-07-28
19GPT-5.6 Terra OpenAI35%maxEpoch AI2026-07-28
20DeepSeek V4 Flash DeepSeek34%maxEpoch AI2026-08-05
21Gemini 3.1 Pro Preview Google34%highEpoch AI2026-07-27
22Grok 4.6 xAI34%xhighEpoch AI2026-08-14
23GLM-5.3 Z.ai (Zhipu)33%maxEpoch AI2026-08-30
24Gemini 3.5 Flash Google32%highEpoch AI2026-07-27
25Qwen3.7 Max Alibaba (Qwen)32%Epoch AI2026-07-28
26Claude Haiku 5.5 Anthropic30%maxEpoch AI2026-10-07
27Gemini 3.6 Flash Google30%highEpoch AI2026-08-05
28Grok 4.7 xAI29%xhighEpoch AI2026-09-22
29o3 OpenAI29%highEpoch AI2026-08-27
30Claude Opus 4.7 Anthropic28%maxEpoch AI2026-07-25
31Gemini 3 Flash Preview Google26%lowEpoch AI2026-08-05
32Kimi K3 Moonshot AI26%maxEpoch AI2026-07-29
33Claude Opus 4.6 Anthropic25%maxEpoch AI2026-07-25
34Muse Spark 1.3 Meta25%maxEpoch AI2026-09-18
35GPT-5 OpenAI23%highEpoch AI2026-08-05
36GPT-5.2 OpenAI23%highEpoch AI2026-08-06
37Claude Opus 4.5 Anthropic22%48KEpoch AI2026-07-25
38Qwen3.6 35B-A3B Alibaba (Qwen)22%noneEpoch AI2026-08-27
39Claude Opus 4.1 Anthropic21%24KEpoch AI2026-07-25
40GPT-5.6 Luna OpenAI21%maxEpoch AI2026-07-28
41Nemotron 3 Ultra NVIDIA20%Epoch AI2026-08-27
42Qwen3.5-Flash Alibaba (Qwen)20%Epoch AI2026-08-27
43Gemini 3.5 Flash Lite Google19%lowEpoch AI2026-08-05
44GLM-5.2 Z.ai (Zhipu)19%lowEpoch AI2026-08-27
45GPT-5.1 OpenAI19%lowEpoch AI2026-08-27
46Qwen3.6 Max Preview Alibaba (Qwen)19%noneEpoch AI2026-08-27
47Kimi K2.6 Moonshot AI18%Epoch AI2026-07-17
48Qwen3.5 397B-A17B Alibaba (Qwen)18%noneEpoch AI2026-08-27
49Qwen3.6 Flash Alibaba (Qwen)18%noneEpoch AI2026-08-27
50Claude Sonnet 4.5 Anthropic17%48KEpoch AI2026-07-25
51Qwen3.5 122B-A10B Alibaba (Qwen)17%noneEpoch AI2026-08-27
52Qwen3.5 Plus Alibaba (Qwen)17%noneEpoch AI2026-08-27
53Qwen3.7 Plus Alibaba (Qwen)17%Epoch AI2026-08-28
54Claude Sonnet 4.6 Anthropic16%lowEpoch AI2026-08-06
55Qwen3.7 Flash Alibaba (Qwen)15%Epoch AI2026-08-28
56GPT-4 OpenAI12%Epoch AI2026-08-28
57GPT-4o mini OpenAI12%Epoch AI2026-08-27
58Qwen3.6 Plus Alibaba (Qwen)12%noneEpoch AI2026-08-27
59GPT-5.4 mini OpenAI11%noneEpoch AI2026-08-27
60GPT-5 Mini OpenAI10%minimalEpoch AI2026-08-27
61GPT-5.4 nano OpenAI9%noneEpoch AI2026-08-27
62GPT-5 Nano OpenAI9%lowEpoch AI2026-08-27
63Qwen3 235B-A22B Alibaba (Qwen)9%Epoch AI2026-08-27
64GLM-5.3-Flash Z.ai (Zhipu)8%maxEpoch AI2026-08-28
65MiniMax-M3 MiniMax8%noneEpoch AI2026-08-27
66GPT-4.1 mini OpenAI7%Epoch AI2026-08-27
67GPT-6 Luna OpenAI7%maxEpoch AI2026-09-22
68o3-mini OpenAI7%highEpoch AI2026-08-27
69Qwen3.6 27B Alibaba (Qwen)7%noneEpoch AI2026-08-27
70Inkling-Small Thinking Machines Lab6%xhighEpoch AI2026-08-15
71o4-mini OpenAI5%highEpoch AI2026-08-27
72Qwen3 Max Alibaba (Qwen)5%Epoch AI2026-08-27
73GPT-3.5-turbo OpenAI3%Epoch AI2026-08-27
74gpt-oss-120b OpenAI2%mediumEpoch AI2026-08-27

Compare the leaders

Other reasoning benchmarks

Frequently asked questions

Which model has the highest Mystery Game Puzzles score?

As of October 2026, GPT-6 Astra has the highest published Mystery Game Puzzles score on Noometry at 84%, out of 74 models with results.

What is the best open-weight model on Mystery Game Puzzles?

DeepSeek V4.1 Flash has the highest Mystery Game Puzzles accuracy among open-weight models at 43%, ranking 12 of 74 overall.