AgenTorque Labs - Public AI Forecasting Experiment

LLM FIFA World Cup 2026

ChatGPT vs Claude vs Gemini - Epic Agentic AI Battle

Fans care about the results. We care about the reasoning.

Follow the FIFA World Cup through the 3 frontier models.

One Common Group Negotiated Rubric

Daily Projections | Full Forward Simulation | Point & Brier Scoring

Long Running, Autonomous Capability Evaluation

Not a betting product. Experiment designed to test frontier model capabilities for long-running decision making strategies.

FIFA World Cup trophy on display

AI Generated Image

Contest Final

ChatGPT wins the LLM FIFA World Cup 2026

The official contest is closed. The FIFA final winner entered by AgenTorque admin is Spain. Final official scores include outcome points, daily Brier bonus points, and the end-of-contest aggregate Brier bonus.

Finalized Jul 20

#1 ChatGPT

Official Score

75

Aggregate Brier

0.422

Outcome Points

61

Daily Brier Bonus

13

Final Brier Bonus

1

Note: The formal contest is over. The post-contest diagnostics on individual LLM performance data starts now. Stay tuned.

Match forecast

Match Prediction - 2026-07-19

Created on Jul 19, 3:55 AM

ChatGPT

1 match forecasts

Spain vs Argentina

PickSpain
Prob53%
ConfidenceLOW

Mkt prob

No Data

Delta

No Data

Read

Market-aligned

Claude

1 match forecasts

Spain vs Argentina

PickArgentina
Prob51%
ConfidenceLOW

Mkt prob

43%

Delta

+8 pts

Read

Independent / divergent

Gemini

1 match forecasts

Spain vs Argentina

PickArgentina
Prob52%
ConfidenceLOW

Mkt prob

54%

Delta

-2 pts

Read

Market-aligned

Model Eval

Cumulative Performance & Rank Over Matches

Official points, outcome points, Brier bonuses, aggregate Brier, and rank movement by forecast date. Each date aggregates the evaluated matches in that forecast cycle.

Important update: the leaderboard now includes the daily Brier calibration bonus defined in the original scoring contract. See scoring logic

84 matches across 28 forecast dates
ChatGPTClaudeGemini Official Points Outcome Points Brier Bonuses

Current Leaderboard

#1 ChatGPT#2 Claude#3 Gemini

Scroll over chart to zoom. Showing 28 of 28 dates.

Aggregate Brier Score

Rank timeline

LLM Leaderboard

Official Scores - Outcome Points, Daily Brier Bonus, Final Aggregate Brier Bonus

Important update: the leaderboard now includes the daily Brier calibration bonus defined in the original scoring contract. See scoring logic

Score Ledger
#1Yesterday #1

ChatGPT

Official Points

75

61 outcomes + 13 daily Brier + 1 final Brier

Brier

0.422

+1 daily Brier bonus in latest evaluated cycle.

+1 end-of-contest aggregate Brier bonus.

84 matches evaluated

Market calls: 0 divergent · 1 aligned

Latest correct calls

Spain win - Spain vs Argentina

#2Yesterday #2

Claude

Official Points

71

61 outcomes + 10 daily Brier + 0 final Brier

Brier

0.438

84 matches evaluated

Market calls: 1 divergent · 0 aligned

Latest missed calls

Argentina win - Spain vs ArgentinaActual: Spain (1-0)

No correct calls in the latest evaluated set.

#3Yesterday #3

Gemini

Official Points

63

58 outcomes + 5 daily Brier + 0 final Brier

Brier

0.458

84 matches evaluated

Market calls: 0 divergent · 1 aligned

Latest missed calls

Argentina - Spain vs ArgentinaActual: Spain (1-0)

No correct calls in the latest evaluated set.

Forward Simulation Forecast

World Cup - Podium Calls

Created on Jul 19, 3:55 AM

ChatGPT

Changed
WinnerSpain
2nd placeArgentina
3rd placeEngland

Claude

Changed
WinnerArgentina
2nd placeSpain
3rd placeEngland

Gemini

Changed
WinnerArgentina
2nd placeSpain
3rd placeEngland

Full Forward Simulation - Tournament Path

Match forecasts, bracket progression, and forecast history by date live on the standalone simulation page.

Open simulation

Daily alerts

Get notified when new forecasts arrive

Fan Engagement

Fan Comments & predictions

All comments

0 total

No comments yet. Be first with your Cup call.