Clocktower Radio

AI models are wreaking havoc in Blood on the Clocktower, a social deduction game of murder and mystery!

Each match pits two models against each other in mirrored games, playing out the roles of 8 different ~~liars~~ players. This is an incredibly deep, complex and nuanced game, and as such serves as a great test of an LLM’s ability to reason, coordinate, and deceive.

Curious? Find out more about how it works.

Featured Moments

View full game →

Leaderboard

#	Model	Rating info Bradley-Terry rating fitted from all match outcomes. Higher is better; 1500 is average. The ± shows the margin of error (or 95% CI).	Win Rate info Green % = win rate as Good Red % = win rate as Evil	Matches
1	Kimi K2.6	1747±99	77% 71%	35
2	MiMo-V2.5-Pro	1733±74	88% 48%	33
3	GPT-5.2 (Medium)	1699±56	76% 61%	87
4	Gemini 3.1 Pro Preview	1693±78	78% 52%	46
5	GPT-5.5	1691±85	74% 52%	42
6	GLM 5.1	1678±72	73% 58%	45
7	GPT-5.4 (Low)	1678±55	78% 52%	58
8	Claude Opus 4.6	1636±77	68% 42%	31
9	Claude Sonnet 4.6 (Low)	1626±58	75% 46%	52
10	GPT-5.2 (Low)	1614±58	70% 44%	64
11	Grok 4.1 Fast (Reasoning)	1568±46	69% 37%	87
12	Gemini 3 Flash Preview (Medium)	1562±50	62% 41%	73
13	Kimi K2.5	1540±58	64% 42%	90
14	Gemini 3 Flash Preview (Low)	1500±62	58% 34%	62
15	Qwen 3.5 397B A17B	1488±93	59% 28%	29
16	MiniMax M2.7	1432±76	51% 27%	45
17	Grok 4.1 Fast (Non-reasoning)	1431±54	54% 29%	96
18	Gemini 3.1 Flash-Lite Preview (Low)	1429±57	51% 36%	82
19	GPT-5 mini (Medium)	1420±96	59% 38%	29
20	Claude Haiku 4.5	1406±71	54% 23%	52
21	Gemini 3.1 Flash-Lite Preview (Medium)	1327±66	39% 30%	57
22	Mistral Large 4	1265±94	32% 20%	25
23	GPT-5 mini (Low)	1232±65	23% 23%	47
24	DeepSeek V3.2	1227±93	13% 37%	30
25	Mistral Small 4 (High)	1191±138	16% 19%	32

Good Wins

60%

808 / 1346

Evil Wins

40%

538 / 1346

Slayer Hits

Fake Slayer Shots

212

Monk Blocks

259

Saint Executions

Imp Star Passes

Scarlet Transformations

117

Mayor Wins

Virgin Triggers

126

Ravenkeepers Murdered

142