ForecastBench Tournament Overall Brier Index
Forecasting · 2024-09-30
Equal-weighted Brier Index across ForecastBench tournament questions.
Top models (higher is better)
| Model | Score |
|---|---|
| Grok 4.20 | 67.9 |
| GPT-5 | 65.9 |
| Gemini 3 Pro Preview | 65.7 |
| Grok 4 | 65.5 |
| GPT-4.5 | 65.4 |
| Claude 3.7 Sonnet | 65.4 |
| Grok 4 Fast | 65.3 |
| Opus 4.1 | 65.1 |
| Opus 4.5 | 65.1 |
| Grok 4.1 Fast | 65.1 |
| Gemini 3.1 Pro Preview | 65.0 |
| o3 | 65.0 |
| Gemini 2.5 Pro Preview 03-25 | 64.9 |
| Gemini 3 Flash Preview | 64.9 |
| Opus 4 | 64.8 |
| Sonnet 4.6 | 64.7 |
| Gemini 2.5 Pro | 64.7 |
| GPT-5.5 | 64.7 |
| O4 Mini | 64.7 |
| Sonnet 4 | 64.6 |
| Sonnet 4.5 | 64.6 |
| GPT-5.2 | 64.6 |
| Qwen3-235B-A22B | 64.6 |
| GPT-4.1 | 64.4 |
| Grok Beta | 64.4 |
| o3-mini | 64.3 |
| Claude 3.5 Sonnet (Oct 2024) | 64.2 |
| Haiku 4.5 | 64.2 |
| GPT-5 Mini | 64.1 |
| Opus 4.7 | 64.0 |
| Gemini 2.5 Flash Preview 04-17 | 64.0 |
| GLM-5 | 64.0 |
| DeepSeek-R1 | 63.9 |
| Gemini 3.5 Flash | 63.9 |
| GPT-5.4 | 63.8 |
| Opus 4.6 | 63.7 |
| Grok 4.3 | 63.7 |
| Claude 3.5 Sonnet (June 2024) | 63.6 |
| GPT-4 Turbo | 63.6 |
| Gemini 2.5 Flash | 63.3 |