Mystery Game Puzzles (Epoch AI) (Mystery Game Puzzles)
We show this table for reference; we do not rank on it.
Epoch AI's Mystery Game Puzzles benchmark asks models for the single best next move in 100 positions from a puzzle variant of a well-known game whose identity Epoch keeps secret.
Benchmark score on Mystery Game Puzzles — November 13, 2026
We compile the Mystery Game Puzzles rows from benchmark-owner or independent runs. Claude Fable 5.2 leads the table at 90.0%, followed by GPT-6 Astra (84.0%) and GPT-6.1 Sol (80.0%). We do not use these results to rank models overall.
Claude Fable 5.2
Anthropic
Run by Mystery Game PuzzlesGPT-6 Astra
OpenAI
Run by Mystery Game PuzzlesGPT-6.1 Sol
OpenAI
Run by Mystery Game Puzzles79 modelsReasoningCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (79 models)
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.2AnthropicRun by Mystery Game Puzzles | 90.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 2 | GPT-6 AstraOpenAIRun by Mystery Game Puzzles | 84.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 3 | GPT-6.1 SolOpenAIRun by Mystery Game Puzzles | 80.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 4 | GPT-6.1 TerraOpenAIRun by Mystery Game Puzzles | 74.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 5 | Gemini 4 CarbonGoogleRun by Mystery Game Puzzles | 73.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 6 | Claude Opus 5.5AnthropicRun by Mystery Game Puzzles | 71.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 7 | Claude Sonnet 5.5AnthropicRun by Mystery Game Puzzles | 65.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 8 | Gemini 4 ArgonGoogleRun by Mystery Game Puzzles | 64.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 9 | Claude Opus 5AnthropicRun by Mystery Game Puzzles | 59.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 10 | Claude Fable 5.1AnthropicRun by Mystery Game Puzzles | 58.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 11 | GPT-5.6 SolOpenAIRun by Mystery Game Puzzles | 58.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 12 | GPT-5.5OpenAIRun by Mystery Game Puzzles | 56.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 13 | GPT-6 SolOpenAIRun by Mystery Game Puzzles | 56.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 14 | Claude Fable 5AnthropicRun by Mystery Game Puzzles | 52.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 15 | Muse WonderMetaRun by Mystery Game Puzzles | 51.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 16 | DeepSeek V4.1 Flash 1017DeepSeekRun by Mystery Game Puzzles | 49.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 17 | Gemini 3.8 FlashGoogleRun by Mystery Game Puzzles | 47.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 18 | DeepSeek V4 Pro 0813DeepSeekRun by Mystery Game Puzzles | 43.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 19 | Kimi K3.1 ProMoonshot AIRun by Mystery Game Puzzles | 42.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 20 | Qwen3.8 MaxAlibabaRun by Mystery Game Puzzles | 38.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 21 | GPT-5.4OpenAIRun by Mystery Game Puzzles | 37.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 22 | Gemini 3.7 FlashGoogleRun by Mystery Game Puzzles | 37.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 23 | Claude Opus 4.8AnthropicRun by Mystery Game Puzzles | 36.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 24 | Claude Sonnet 5AnthropicRun by Mystery Game Puzzles | 35.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 25 | GPT-5.6 TerraOpenAIRun by Mystery Game Puzzles | 35.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 26 | Gemini 3.1 ProGoogleRun by Mystery Game Puzzles | 34.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 27 | Grok 4.6xAIRun by Mystery Game Puzzles | 34.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 28 | DeepSeek V4 Flash 0731DeepSeekRun by Mystery Game Puzzles | 34.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 29 | GLM-5.3Z.AIRun by Mystery Game Puzzles | 33.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 30 | Qwen3.7 MaxAlibabaRun by Mystery Game Puzzles | 32.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 31 | Gemini 3.5 FlashGoogleRun by Mystery Game Puzzles | 32.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 32 | Claude Haiku 5.5AnthropicRun by Mystery Game Puzzles | 30.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 33 | Gemini 3.6 FlashGoogleRun by Mystery Game Puzzles | 30.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 34 | o3OpenAIRun by Mystery Game Puzzles | 29.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 35 | Grok 4.7xAIRun by Mystery Game Puzzles | 29.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 36 | Claude Opus 4.7AnthropicRun by Mystery Game Puzzles | 28.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 37 | Kimi K3Moonshot AIRun by Mystery Game Puzzles | 26.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 38 | Claude Opus 4.6AnthropicRun by Mystery Game Puzzles | 25.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 39 | Muse Spark 1.3MetaRun by Mystery Game Puzzles | 25.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 40 | GPT-5.2OpenAIRun by Mystery Game Puzzles | 23.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 41 | GPT-5 (high)OpenAIRun by Mystery Game Puzzles | 23.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 42 | Claude Opus 4.5AnthropicRun by Mystery Game Puzzles | 22.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 43 | Qwen3.6-35B-A3BAlibabaRun by Mystery Game Puzzles | 22.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 44 | GPT-5.6 LunaOpenAIRun by Mystery Game Puzzles | 21.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 45 | Nemotron 3 UltraNVIDIARun by Mystery Game Puzzles | 20.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 46 | Gemini 3 FlashGoogleRun by Mystery Game Puzzles | 20.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 47 | Qwen3.5 FlashAlibabaRun by Mystery Game Puzzles | 20.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 48 | Gemini 3.5 Flash-LiteGoogleRun by Mystery Game Puzzles | 19.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 49 | Qwen 3.6 Max (preview)AlibabaRun by Mystery Game Puzzles | 19.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 50 | Kimi K2.6Moonshot AIRun by Mystery Game Puzzles | 18.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 51 | Qwen3.5 397BAlibabaRun by Mystery Game Puzzles | 18.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 52 | Qwen3.7 PlusAlibabaRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 53 | Qwen3.5-122B-A10BAlibabaRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 54 | Mistral Large 4MistralRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 55 | Claude Sonnet 4.5AnthropicRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 56 | Qwen3.5 PlusAlibabaRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 57 | DeepSeek V4 ProDeepSeekRun by Mystery Game Puzzles | 17.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 58 | GPT-5.1OpenAIRun by Mystery Game Puzzles | 16.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 59 | Claude Sonnet 4.6AnthropicRun by Mystery Game Puzzles | 16.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 60 | GLM-5.2Z.AIRun by Mystery Game Puzzles | 15.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 61 | GPT-5 (medium)OpenAIRun by Mystery Game Puzzles | 15.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 62 | Qwen3.7 FlashAlibabaRun by Mystery Game Puzzles | 15.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 63 | Qwen3.6 PlusAlibabaRun by Mystery Game Puzzles | 12.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 64 | GPT-4o miniOpenAIRun by Mystery Game Puzzles | 12.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 65 | Qwen3 235B 2507 (Reasoning)AlibabaRun by Mystery Game Puzzles | 9.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 66 | MiniMax M3MiniMaxRun by Mystery Game Puzzles | 8.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 67 | GLM-5.3-FlashZ.AIRun by Mystery Game Puzzles | 8.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 68 | GPT-5 nanoOpenAIRun by Mystery Game Puzzles | 8.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 69 | GPT-5.4 miniOpenAIRun by Mystery Game Puzzles | 7.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 70 | Qwen3.6-27BAlibabaRun by Mystery Game Puzzles | 7.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 71 | GPT-6 LunaOpenAIRun by Mystery Game Puzzles | 7.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 72 | o3-miniOpenAIRun by Mystery Game Puzzles | 7.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 73 | GPT-4.1 miniOpenAIRun by Mystery Game Puzzles | 7.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 74 | Inkling-SmallThinking Machines LabRun by Mystery Game Puzzles | 6.0% | Run by Mystery Game Puzzles | Not reported | Open |
| 75 | GPT-5.4 nanoOpenAIRun by Mystery Game Puzzles | 5.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 76 | o4-mini (high)OpenAIRun by Mystery Game Puzzles | 5.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 77 | Qwen3 MaxAlibabaRun by Mystery Game Puzzles | 5.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 78 | GPT-5 miniOpenAIRun by Mystery Game Puzzles | 5.0% | Run by Mystery Game Puzzles | Not reported | Closed |
| 79 | GPT-OSS 120BOpenAIRun by Mystery Game Puzzles | 0.0% | Run by Mystery Game Puzzles | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
How to read this leaderboard
Higher is better. With 100 positions, one answer is one point, so gaps of a few points are within noise. Claude Fable 5.2 leads at 90%.
Operator receipt: 79 sourced rows are currently displayable on this page; the leading published row is Claude Fable 5.2 at 90.0%.
Honest limit: Because the game, prompt and transcripts are secret, outside readers cannot inspect the task or check individual answers. Like Chess Puzzles it measures a narrow puzzle skill. BenchLM shows it as display-only; it does not enter overall or category rankings.
Among the reported Mystery Game Puzzles rows, Claude Fable 5.2 is first at 90.0%. The third row is 10.0 points behind. The broader top-10 range is 32.0 points, so the table still separates the published systems.
79 models have been evaluated on Mystery Game Puzzles. The benchmark falls in the Reasoning category. We keep these results separate from the weighted global scoring system, so they remain source-specific evidence. Mystery Game Puzzles is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About Mystery Game Puzzles
Year
2026
Tasks
100 single-best-move positions from an undisclosed game
Format
Exact-match move answer from text input, minimal agent scaffold
Difficulty
Unfamiliar game positions, single best move
The setup mirrors Epoch's Chess Puzzles: positions come from random games rolled forward to mid-game and filtered to states with one best move. Models receive the state as text in a minimal agent scaffold and submit a move through a tool. Answers are normalized to the game's standard notation and must match the key; an illegal move is wrong and an answer with no valid move gets no credit. Epoch does not publish the prompt, examples or transcripts, to keep the game hidden and make benchmark-specific preparation harder.
Freshness and provenance
Version
Mystery Game Puzzles 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
Why is the game secret?
Epoch keeps the game's identity undisclosed so labs cannot prepare for the benchmark specifically. It is otherwise built like Epoch's Chess Puzzles.
Which model leads Mystery Game Puzzles?
Claude Fable 5.2 at max effort leads with 90%, ahead of GPT-6 Astra (84%), GPT-6.1 Sol (80%) and GPT-6.1 Terra (74%).
Compare top models on Mystery Game Puzzles
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.