Benchmark Almanac

MafiaBench

openai/chatgpt-4o-latest 1,740 tops its leaderboard Gemini 2.0 Flash Feb 2025 1,561 leads the models we track
Not in model cards

Models play a social deduction game with hidden roles and public discussion.

The public leaderboard compares at least six model systems. GPT-4o-mini is listed at 36% win rate in the checked page.

Made by
MafiaBench
Released
April 2025
Kind
Multi-agent games
Level
Everyday
Scored
Settled by events
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Actions, Text
Interaction
Multi-agent
Environment
Mafia
Items from
Live play or votes
Refresh
Live
Run by
A host
In views
Games, Agents, Kept fresh
Strongest model tested
Claude 3.7 Sonnet #119 on the capability index
Last model added
February 2025
Scores refreshed
November 20th, 2026
6
Entries on its own leaderboard, 4 models we track

openai/chatgpt-4o-latest tops the leaderboard with 1,740 Gemini 2.0 Flash leads with 1,561 among models released by February 2025

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Gemini 2.0 Flash, at 1,561. This view shows only models we track, each at its best.

The newest model with a score here came out in February 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

Bars start at 1,300.

  1. 1 openai/chatgpt-4o-latest 1,740
  2. 2 google/gemini-pro-1.5 1,627
  3. 3 Gemini 2.0 Flash Feb 2025 1,561
  4. 4 Claude 3.5 Sonnet 1,421
  5. 5 Claude 3.7 Sonnet 1,349
  6. 6 GPT-4o mini 1,302

Source: Official MafiaBench leaderboard

Source: Official MafiaBench leaderboard

#EntryScoreSettingDateSource
1 openai/chatgpt-4o-latest 1,740 Official leaderboard
2 google/gemini-pro-1.5 1,627 Official leaderboard
3 Gemini 2.0 Flash Feb 2025 1,561 Feb 2025 Official leaderboard
4 Claude 3.5 Sonnet 1,421 Jun 2024 Official leaderboard
5 Claude 3.7 Sonnet 1,349 Feb 2025 Official leaderboard
6 GPT-4o mini 1,302 Jul 2024 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 2.0 Flash Feb 2025 1,561 Feb 2025 Official leaderboard
2 Claude 3.5 Sonnet 1,421 Jun 2024 Official leaderboard
3 Claude 3.7 Sonnet 1,349 Feb 2025 Official leaderboard
4 GPT-4o mini 1,302 Jul 2024 Official leaderboard

Rankings by part

We read 2 parts from the benchmark’s own leaderboard: Win Rate As Mafia and Win Rate As Town. Each ranks models on one part of the test. Claude 3.5 Sonnet leads Win Rate As Mafia and GPT-4o mini leads Win Rate As Town. Arrows show how far a model moved from its overall rank among the models we track.

Win Rate As Mafia

4 models we track have a score here.

  1. 1 Claude 3.5 Sonnet ↑1 49%
  2. 2 Gemini 2.0 Flash Feb 2025 ↓1 45%
  3. 3 Claude 3.7 Sonnet 40%
  4. 4 GPT-4o mini 36%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Google DeepMind
Own leaderboard Not tested since Apr 2025

Source: Official MafiaBench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
1,300 1,400 1,500 1,600 2024 2025 1,300 1,400 1,500 1,600 2024 2025

Source: Official MafiaBench leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Games and long tasks

All 261 in Games and long tasks
  • ARC-AGI-3 In use Interactive game environments where agents must explore, infer goals and rules, and act efficiently without instructions. GPT-6.2 Astra 100%
  • Orak Fading LLM agents play 12 popular video games across most genres through an MCP interface. Each game scores in-game progress or outcomes, and the board ranks models by their average rank across the games. Gemini 2.5 Pro 3.90
  • MineDojo Fading Agents complete open-ended tasks in Minecraft, such as mining blocks and fighting mobs, from a large simulation suite. Programmatic checks score many tasks. Seed2.0 Pro 49%
  • TextQuests Only in system cards Agents play 25 classic interactive fiction games. The score is harmful actions per 100 steps, where lower is better. The board also reports game progress. Llama 4 Scout 7.00
  • GameWorld Only in system cards Multimodal game agents play 34 browser games across 170 tasks. Outcomes are checked from the game state. GPT-6.1 Terra 76.8%
  • Swarm Bench Only in system cards Models orchestrate agent swarms on complex tasks that benefit from coordinated decomposition and parallel execution. Kimi K3 76.3%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 21 months ago, in February 2025, for Claude 3.7 Sonnet. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (6 entries).
  • Pass Models scored 4 models we track, plus 2 other entries on its leaderboard.
  • Warning Newest score The newest came 21 months ago, in February 2025, for Claude 3.7 Sonnet.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude 3.7 Sonnet (#119).
  • Note Headroom Scored as an Elo rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on mafiabench.org: 4 rows for models we track and 2 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check