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	1811±118	86% 79%	28
2	GPT-5.2 (Medium)	1715±65	77% 65%	77
3	GLM 5.1	1700±73	75% 62%	40
4	Gemini 3.1 Pro Preview	1700±93	85% 56%	34
5	GPT-5.4 (Low)	1694±57	78% 53%	55
6	Claude Opus 4.6	1639±79	65% 46%	26
7	GPT-5.2 (Low)	1633±59	71% 45%	62
8	Claude Sonnet 4.6 (Low)	1632±55	77% 47%	47
9	Grok 4.1 Fast (Reasoning)	1590±47	71% 38%	85
10	Gemini 3 Flash Preview (Medium)	1577±52	63% 41%	70
11	Kimi K2.5	1563±58	67% 44%	87
12	Gemini 3 Flash Preview (Low)	1519±62	61% 34%	59
13	Qwen 3.5 397B A17B	1501±95	59% 28%	29
14	MiniMax M2.7	1457±76	55% 29%	42
15	Gemini 3.1 Flash-Lite Preview (Low)	1441±58	52% 37%	78
16	Grok 4.1 Fast (Non-reasoning)	1439±53	54% 29%	92
17	GPT-5 mini (Medium)	1423±97	57% 39%	28
18	Claude Haiku 4.5	1417±73	53% 24%	51
19	Gemini 3.1 Flash-Lite Preview (Medium)	1342±66	39% 30%	57
20	Mistral Large 4	1285±95	35% 22%	23
21	GPT-5 mini (Low)	1248±68	24% 24%	45
22	DeepSeek V3.2	1241±94	13% 37%	30
23	Mistral Small 4 (High)	1209±133	17% 20%	30

Good Wins

60%

712 / 1192

Evil Wins

40%

480 / 1192

Slayer Hits

Fake Slayer Shots

199

Monk Blocks

228

Saint Executions

Imp Star Passes

Scarlet Transformations

108

Mayor Wins

Virgin Triggers

105

Ravenkeepers Murdered

125