General
Mystery Game Puzzles leaderboard
As of 2026-11-18, the highest measured score on Mystery Game Puzzles is 95.0% by GPT-6.2 Astra. 70 more models have estimated scores, calibrated from the benchmarks they were measured on.
Measured scores: benchlm.ai.
| # | Model | Score | Source |
|---|---|---|---|
| 1 | GPT-6.2 Astra | 95.0% | measured |
| 2 | Claude Fable 5.2 | 90.0% | measured |
| 3 | GPT-6.2 Sol | 87.5% | estimated ± 6.4 pp, low confidence |
| 4 | GPT-6 Cyber | 86.3% | estimated ± 6.4 pp, low confidence |
| 5 | GPT-6 Astra | 84.0% | measured |
| 6 | GPT-6.1 Sol | 80.0% | measured |
| 7 | Claude Mythos 5.2 | 77.5% | estimated ± 7.5 pp, medium confidence |
| 8 | GPT-6.1 Terra | 74.0% | measured |
| 9 | Gemini 4 Carbon | 73.0% | measured |
| 10 | Claude Opus 5.5 | 71.0% | measured |
| 11 | Claude Sonnet 5.5 | 65.0% | measured |
| 12 | Gemini 4 Argon | 64.0% | measured |
| 13 | Hy4 | 63.6% | estimated ± 12.6 pp, low confidence |
| 14 | ERNIE 5.5 Flash Preview | 61.0% | estimated ± 12.6 pp, low confidence |
| 15 | Claude Opus 5 | 59.0% | measured |
| 16 | GPT-5.5 Pro | 58.7% | estimated ± 11.3 pp, low confidence |
| 17 | Claude Fable 5.1 | 58.0% | measured |
| 18 | GPT-5.6 Sol | 58.0% | measured |
| 19 | GPT-5.5 | 56.0% | measured |
| 20 | GPT-6 Sol | 56.0% | measured |
| 21 | Claude Mythos 5 | 54.8% | estimated ± 7.5 pp, medium confidence |
| 22 | GPT-5.4 Pro | 53.7% | estimated ± 11.3 pp, low confidence |
| 23 | Claude Fable 5 | 52.0% | measured |
| 24 | Muse Wonder | 51.0% | measured |
| 25 | Claude Opus 4.6 (Adaptive) | 49.1% | estimated ± 14.6 pp, low confidence |
| 26 | DeepSeek V4.1 Flash 1017 | 49.0% | measured |
| 27 | Gemini 3.8 Flash | 47.0% | measured |
| 28 | Grok 4.20 Multi-agent | 45.3% | estimated ± 14.6 pp, low confidence |
| 29 | Claude Mythos Preview | 45.1% | estimated ± 7.5 pp, medium confidence |
| 30 | Grok 4.1 Fast (Reasoning) | 42.0% | estimated ± 14.6 pp, low confidence |
| 31 | Kimi K3.1 Pro | 42.0% | measured |
| 32 | Qwen3.8 Max | 38.0% | measured |
| 33 | Gemini 3.7 Flash | 37.0% | measured |
| 34 | GPT-5.4 | 37.0% | measured |
| 35 | Claude Opus 4.8 | 36.0% | measured |
| 36 | Claude Sonnet 5 | 35.0% | measured |
| 37 | GPT-5.6 Terra | 35.0% | measured |
| 38 | DeepSeek V4 Flash 0731 | 34.0% | measured |
| 39 | Gemini 3.1 Pro | 34.0% | measured |
| 40 | Grok 4.6 | 34.0% | measured |
| 41 | GLM-5.3 | 33.0% | measured |
| 42 | Gemini 3.5 Flash | 32.0% | measured |
| 43 | Qwen3.7 Max | 32.0% | measured |
| 44 | MiniMax M3.1 Flash | 31.8% | estimated ± 13.3 pp, low confidence |
| 45 | GLM-5.1 | 30.8% | estimated ± 9.8 pp, medium confidence |
| 46 | Muse Spark Open | 30.4% | estimated ± 13.3 pp, low confidence |
| 47 | Hy4 preview | 30.2% | estimated ± 13.3 pp, low confidence |
| 48 | Claude Haiku 5.5 | 30.0% | measured |
| 49 | Gemini 3.6 Flash | 30.0% | measured |
| 50 | DeepSeek V4 Pro | 29.0% | estimated ± 13.3 pp, low confidence |
| 51 | Grok 4.7 | 29.0% | measured |
| 52 | o3 | 29.0% | measured |
| 53 | Ling 3.1 Flash | 28.6% | estimated ± 13.3 pp, low confidence |
| 54 | Claude Opus 4.7 | 28.0% | measured |
| 55 | Grok 4.5 | 28.0% | estimated ± 10.7 pp, low confidence |
| 56 | Hy3 | 27.0% | estimated ± 9.8 pp, medium confidence |
| 57 | MiMo-V2.6-Pro | 27.0% | estimated ± 7.5 pp, medium confidence |
| 58 | Kimi K3 | 26.0% | measured |
| 59 | Qwen3.8-27B | 25.9% | estimated ± 12.4 pp, low confidence |
| 60 | Step 5 Preview | 25.3% | estimated ± 13.3 pp, low confidence |
| 61 | Claude Opus 4.6 | 25.0% | measured |
| 62 | Muse Spark 1.3 | 25.0% | measured |
| 63 | Muse Spark 1.1 | 23.5% | estimated ± 11.3 pp, low confidence |
| 64 | GPT-5 | 23.0% | measured |
| 65 | GPT-5.2 | 23.0% | measured |
| 66 | Claude Opus 4.5 | 22.0% | measured |
| 67 | Qwen3.6-35B-A3B | 22.0% | measured |
| 68 | GPT-5.6 Luna | 21.0% | measured |
| 69 | Grok 4.20 | 20.2% | estimated ± 11.3 pp, low confidence |
| 70 | Gemini 3 Flash | 20.0% | measured |
| 71 | Nemotron 3 Ultra | 20.0% | measured |
| 72 | Qwen3.5 Flash | 20.0% | measured |
| 73 | Beam 1.0 501B-A23B | 19.1% | estimated ± 13.3 pp, low confidence |
| 74 | Muse Spark 1.2 | 19.1% | estimated ± 11.3 pp, low confidence |
| 75 | Claude Sonnet 4.5 Thinking | 19.1% | estimated ± 4.7 pp, medium confidence |
| 76 | Gemini 3.5 Flash-Lite | 19.0% | measured |
| 77 | Qwen 3.6 Max (preview) | 19.0% | measured |
| 78 | Seed 2.1 Turbo | 18.6% | estimated ± 11.3 pp, low confidence |
| 79 | SWE-2 Flash | 18.0% | estimated ± 7.5 pp, medium confidence |
| 80 | Kimi K2.6 | 18.0% | measured |
| 81 | Qwen3.5 397B | 18.0% | measured |
| 82 | Gemma 4 31B | 17.2% | estimated ± 13.3 pp, low confidence |
| 83 | Claude Sonnet 4.5 | 17.0% | measured |
| 84 | Mistral Large 4 | 17.0% | measured |
| 85 | Qwen3.5-122B-A10B | 17.0% | measured |
| 86 | Qwen3.5 Plus | 17.0% | measured |
| 87 | Qwen3.7 Plus | 17.0% | measured |
| 88 | Gemini 3 Pro | 16.9% | estimated ± 2.2 pp, medium confidence |
| 89 | GPT-5.2-Codex | 16.9% | estimated ± 2.2 pp, medium confidence |
| 90 | Inkling | 16.4% | estimated ± 11.3 pp, low confidence |
| 91 | Seed 2.1 Pro | 16.4% | estimated ± 11.3 pp, low confidence |
| 92 | Claude Sonnet 4.6 | 16.0% | measured |
| 93 | GPT-5.1 | 16.0% | measured |
| 94 | Claude Opus 4.5 Thinking | 15.3% | estimated ± 14.6 pp, low confidence |
| 95 | GLM-4.7 | 15.1% | estimated ± 2.2 pp, low confidence |
| 96 | GLM-5.2 | 15.0% | measured |
| 97 | Qwen3.7 Flash | 15.0% | measured |
| 98 | Grok 4.3 | 14.7% | estimated ± 9.8 pp, medium confidence |
| 99 | Claude Haiku 4.5 Thinking | 14.5% | estimated ± 4.7 pp, medium confidence |
| 100 | Qwen3.5-27B | 14.5% | estimated ± 14.6 pp, low confidence |
| 101 | Kimi K2.7 Code | 14.2% | estimated ± 11.3 pp, low confidence |
| 102 | DeepSeek V3.2 | 14.1% | estimated ± 2.2 pp, low confidence |
| 103 | GPT-5.1-Codex-Max | 14.1% | estimated ± 2.2 pp, low confidence |
| 104 | Step 3.7 Flash | 13.8% | estimated ± 14.6 pp, low confidence |
| 105 | MiMo-V2.5-Pro | 13.6% | estimated ± 14.6 pp, low confidence |
| 106 | MiMo-V2-Pro | 13.5% | estimated ± 14.6 pp, low confidence |
| 107 | Hy3 Preview | 13.5% | estimated ± 14.6 pp, low confidence |
| 108 | Trinity-Large-Thinking | 13.5% | estimated ± 14.6 pp, low confidence |
| 109 | Mistral Medium 3.5 128B | 13.5% | estimated ± 14.6 pp, low confidence |
| 110 | Llama 4 Maverick | 13.5% | estimated ± 14.6 pp, low confidence |
| 111 | Mistral Large 3 | 13.5% | estimated ± 14.6 pp, low confidence |
| 112 | o1 | 13.2% | estimated ± 11.3 pp, low confidence |
| 113 | GPT-4o | 13.0% | estimated ± 11.3 pp, low confidence |
| 114 | MiniMax M2.7 | 12.8% | estimated ± 9.8 pp, low confidence |
| 115 | MiMo-V2.6-Flash | 12.4% | estimated ± 7.5 pp, low confidence |
| 116 | Claude 3 Opus | 12.2% | estimated ± 11.3 pp, low confidence |
| 117 | GPT-4o mini | 12.0% | measured |
| 118 | Qwen3.6 Plus | 12.0% | measured |
| 119 | Gemini 3.1 Flash-Lite | 11.8% | estimated ± 9.8 pp, low confidence |
| 120 | GPT-4.1 nano | 11.7% | estimated ± 11.3 pp, low confidence |
| 121 | Gemma 4 26B A4B | 11.4% | estimated ± 13.3 pp, low confidence |
| 122 | DeepSeek V3.1 | 10.9% | estimated ± 4.7 pp, medium confidence |
| 123 | Grok Code Fast 1 | 10.9% | estimated ± 4.7 pp, medium confidence |
| 124 | Claude Haiku 4.5 | 8.7% | estimated ± 2.2 pp, low confidence |
| 125 | GLM-5.3-Flash | 8.0% | measured |
| 126 | MiniMax M3 | 8.0% | measured |
| 127 | Grok 4 Fast (Reasoning) | 7.3% | estimated ± 4.7 pp, medium confidence |
| 128 | MAI-Thinking-1 | 7.1% | estimated ± 13.3 pp, low confidence |
| 129 | GPT-4.1 mini | 7.0% | measured |
| 130 | GPT-5.4 mini | 7.0% | measured |
| 131 | GPT-6 Luna | 7.0% | measured |
| 132 | o3-mini | 7.0% | measured |
| 133 | Qwen3.6-27B | 7.0% | measured |
| 134 | Grok 4 | 6.9% | estimated ± 2.2 pp, low confidence |
| 135 | Gemini 2.5 Pro | 6.4% | estimated ± 4.7 pp, medium confidence |
| 136 | GPT-4.1 | 6.4% | estimated ± 4.7 pp, medium confidence |
| 137 | Inkling-Small | 6.0% | measured |
| 138 | GPT-5.4 nano | 5.0% | measured |
| 139 | o4-mini | 5.0% | measured |
| 140 | Qwen3 Max | 5.0% | measured |
| 141 | Grok 4.1 Fast | 4.9% | estimated ± 2.2 pp, low confidence |
| 142 | GLM-4.5 | 4.6% | estimated ± 4.7 pp, medium confidence |
| 143 | Gemini 2.5 Flash | 1.0% | estimated ± 4.7 pp, low confidence |
| 144 | GPT-OSS 120B | 0.0% | measured |
All leaderboards
Agentic · terminal
- AA Terminal-Bench 2.1
- AA Terminal-Bench 4.0
- Terminal-Bench 4.0
- Terminal-Bench 2.0
- Terminal-Bench 2.1
- Terminal-Bench 3.0
- Terminal-Bench Hard
- Terminal-Bench-Science 0.1
- Terminal-Bench 4.1
- Terminal-Bench 2.1 (Vals AI)
Agentic · tools
- AA Agentic Index
- AA-AnalystAgent
- AA AutomationBench
- CWE-Bench-AA
- AA Cyber Index
- DeepsecBench-AA
- AA EnterpriseOps-Gym
- GDP.pdf
- AA Harvey LAB v1.0
- Harvey LAB-AA v1.1
- AA ITBench
- AA Tau3 Banking
- Agents' Last Exam
- APEX-Agents
- APEX-Agents-AA
- ApprenticeBench
- AutomationBench
- BFCL v4
- BrowseComp
- Claw-Eval
- CWE-bench v1
- CWE-bench
- CyberGym
- DeepPlanning
- DeepSearchQA
- DRACO
- ExploitGym
- Finance Agent v2
- GDPval-AA
- Gert Labs
- Hermes Bench
- Hermes Index
- HLE w/ tools
- JobBench
- LifeVal
- MazeBench
- MCP Atlas
- MCP-Tasks
- MKA
- OSWorld 2.1
- OSWorld 2.1 Offline
- OSWorld-Verified
- OSWorld 2.0
- PinchBench
- QwenClawBench
- ResearchClawBench
- SkillsBench
- τ²-bench results
- τ³-bench results
- Toolathlon
- Toolathlon-Verified
- VITA-Bench
- WideResearch
Coding
- AA Coding Index
- AA LiveCodeBench
- AA-SciCode
- Bug Hunt Bench
- CursorBench 3.1
- CursorBench 3.2
- CursorBench 4.0
- DeepSWE
- FrontierCode 1.1 Main
- FrontierCode 2.0
- FrontierCode 1.1 Extended
- FrontierSWE v2
- LiveCodeBench
- LiveCodeBench Pro
- LiveCodeBench v6
- NL2Repo
- OpenHarmony Bench v1.2
- OpenHarmony Bench v1.0
- PostTrainBench
- PostTrainBench v1
- PostTrainBench v1.1
- ProgramBench
- React Native Evals
- ScarfBench
- SciCode
- SWE-Atlas Codebase QnA
- SWE-bench Pro
- SWE-bench Verified
- SWE-Rebench
- SWE-Marathon
- Vals LiveCodeBench
- Vals SWE-bench
- Vibe Code Bench
- VulcanBench Frontier v4
- VulcanBench v3
Math
- AIME 2024
- AIME 2025
- AIME26
- FrontierMath v2 (Tier 4)
- FrontierMath v2 (Tiers 1-3)
- HMMT Feb 2025
- HMMT Feb 2026
- HMMT Nov 2025
- IMOAnswerBench
- MATH-500
- MMAnswerBench
Knowledge & reasoning
- AA-GPQA Diamond
- HealthBench Professional (length-adjusted, Artificial Analysis)
- HealthBench Professional (raw, Artificial Analysis)
- AA-HLE
- AA MMLU-Pro
- ARC-AGI-2
- ARC-AGI-3 Provider Adapter
- ARC-AGI-1
- ARC-AGI-3
- Artificial Analysis Intelligence Index
- BBH
- BioBasicsBench
- BioMysteryBench (human-difficult)
- BioMysteryBench (human-solvable)
- C-Eval
- CritPt
- GPQA
- GPQA Diamond
- HealthBench (raw)
- HealthBench Hard
- HealthBench (length-adjusted)
- HealthBench Professional
- HealthBench Professional (raw)
- HLE
- HLE w/o tools
- HLE-Verified
- JevBench 1.4
- LABBench2
- MedXpertQA (Text)
- MMLU
- MMLU-Pro
- MMLU-Pro (Arcee)
- MMLU-Redux
- MMMLU
- AA-Omniscience Accuracy
- SimpleQA
- STEMBench
- SuperGPQA
- Vals GPQA Diamond
- Vals MMLU-Pro
Instruction following
Multilingual
Vision & documents
- AA-MMMU-Pro
- AI2D_TEST
- BabyVision
- BabyVision w/ Python
- BenchCAD Vision2Code (no tools)
- BenchCAD Vision2Code (tools)
- CC-OCR
- Chartography (no tools)
- Chartography (tools)
- CharXiv
- CharXiv w/o tools
- CountBench
- DynaMath
- ERQA
- LVBench
- MathVision
- MathVision w/ Python
- MedXpertQA (MM)
- MLVU (M-Avg)
- MMMU
- MMMU-Pro
- MMMU-Pro w/ Python
- MMVU
- OCRBench V2
- OfficeQA
- OfficeQA Pro
- OmniDocBench 1.5
- ParseBench-1
- ParseBench-2
- PerceptionBench
- RealWorldQA
- RefCOCO (avg)
- ScreenSpot Pro
- SimpleVQA
- SWE-bench Multimodal
- Video-MME (w/o subtitle)
- Video-MME (with subtitle)
- VideoMMMU
- V*
- We-Math
- ZeroBench
Long context
General
- HLE-MultiChoice (Decision Index)
- Artificial Analysis Decision Index v1.0
- Artificial Analysis Decision Index
- When2Call (Decision Index)
- WildGuardTest (Decision Index)
- BabaIsBench
- BinaryAudit
- BlenderWorld (text-only)
- Chess Puzzles
- CompileBench
- Cookie-Bench
- DAYJOB
- DAYJOB Healthcare
- Decision Index Full Score
- EBR-bench
- Epoch Automation Reports
- ExploitBench
- ExploitBench Internal Port
- FACTS Multimodal
- FailureSensorIQ
- FrontierPhysics
- Furniture Assembly
- Decision Index Arts
- Decision Index Knowledge
- Decision Index Language
- Decision Index Retrieval
- Decision Index Tools
- MaxxBench
- MirrorCode
- Mystery Game Puzzles
- NYT Connections
- OTelBench
- SEC-Bench Pro
- SkillsBench official leaderboard (BenchFlow)
- SpireClimb
- StealthMark