ForecastBench Baseline Market Brier Index
Forecasting · 2024-09-30
Difficulty-adjusted Brier Index on baseline market questions.
Top models (higher is better)
| Model | Score |
|---|---|
| o3 | 62.7 |
| MiniMax M3 | 62.0 |
| Kimi K3 | 61.7 |
| GPT-4.5 | 61.4 |
| Opus 4.1 | 61.3 |
| Sonnet 4.6 | 61.3 |
| Claude 3.7 Sonnet | 61.1 |
| Grok Beta | 61.0 |
| O4 Mini | 60.8 |
| Claude 3.5 Sonnet (Oct 2024) | 60.7 |
| Opus 4.8 | 60.5 |
| Sonnet 4.5 | 60.5 |
| Grok 4.20 | 60.5 |
| GPT-4.1 | 60.3 |
| GLM-5 | 60.1 |
| GPT-5.5 | 60.1 |
| Llama 3.3 70B Instruct | 60.1 |
| Haiku 4.5 | 60.0 |
| Grok 4.3 | 60.0 |
| Llama 3 8B Instruct | 59.9 |
| Sonnet 4 | 59.8 |
| Llama 3.1 405B Instruct | 59.8 |
| Gemini 2.5 Pro Preview 03-25 | 59.7 |
| Llama 4 Scout Instruct | 59.5 |
| Opus 4 | 59.4 |
| Sonnet 5 | 59.3 |
| Gemini 2.5 Flash Preview 04-17 | 59.3 |
| Gemini 3 Pro Preview | 59.2 |
| Grok 4.1 Fast | 59.2 |
| Opus 4.7 | 59.0 |
| GPT-4 Turbo | 59.0 |
| GPT-5 | 59.0 |
| Grok 4 | 59.0 |
| Qwen3-235B-A22B | 59.0 |
| Opus 4.5 | 58.9 |
| DeepSeek-R1 | 58.9 |
| GLM 4.5 Air | 58.8 |
| Gemini 1.5 Pro (source unspecified) | 58.6 |
| Gemini 2.5 Pro | 58.6 |
| GPT-5 Mini | 58.3 |