Atlas

Benchmarks

← All benchmarks

ForecastBench Tournament Overall Brier Index

Forecasting · 2024-09-30

Equal-weighted Brier Index across ForecastBench tournament questions.

Top models (higher is better)

ModelScore
Grok 4.2067.4
GPT-5.567.2
Claude Opus 566.5
GPT-5.6 Sol66.5
Opus 4.866.4
Gemini 3.7 Flash66.2
Grok 4.566.1
Gemini 3.6 Flash65.9
GPT-5.6 Sol Pro65.7
Gemini 3.1 Pro Preview65.5
Grok 4.665.5
Gemini 3 Pro Preview65.3
GPT-565.3
Grok 465.0
Opus 4.564.9
GPT-4.564.9
Grok 4 Fast64.9
Claude 3.7 Sonnet64.9
Grok 4.1 Fast64.8
Opus 4.164.7
Gemini 3 Flash Preview64.7
Kimi K364.7
o364.6
Gemini 2.5 Pro Preview 03-2564.5
Gemini 3.5 Flash64.4
Opus 464.3
Sonnet 4.664.3
MiniMax M364.3
Sonnet 464.2
Gemini 2.5 Pro64.2
GPT-5.264.2
O4 Mini64.2
Sonnet 4.564.1
Qwen3-235B-A22B64.1
Grok Beta64.0
GPT-4.163.9
Claude 3.5 Sonnet (Oct 2024)63.8
o3-mini63.8
GLM-563.7
Gemini 2.5 Flash Preview 04-1763.6
Loading Atlas data…