ProphetArena - Agentic: leaderboard
ProphetArena's Agentic scope: models run their own web search in an agentic harness before forecasting live Kalshi prediction-market events, instead of receiving fixed research context. Score is 1 − Brier over the events each predictor forecast (0–1, higher is better); only predictors on the public leaderboard with at least 50 resolved events are ranked.
Metric: 1 - Brier Score. Source: www.prophetarena.co. Saturation forecast: Estimated already saturated. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 0.95 | #6 |
| 2 | Claude Opus 5.5 | 0.95 | #3 |
| 3 | Gemini 3.1 Pro (Preview) | 0.95 | #64 |
| 4 | Claude Fable 5.1 | 0.95 | #10 |
| 5 | Gemini 4 Argon | 0.95 | #9 |
| 6 | Kimi K3 | 0.94 | #35 |
| 7 | Muse Wonder | 0.94 | #18 |
| 8 | GLM-5.2 | 0.94 | #80 |
| 9 | GPT-6.1 Sol | 0.94 | #7 |
| 10 | DeepSeek V4.1 Flash | 0.94 | #37 |
| 11 | GLM-5.3 | 0.94 | #44 |
| 12 | GPT-6 Luna | 0.94 | #61 |
| 13 | Mistral Large 4 | 0.94 | #106 |
| 14 | Gemini 3.8 Flash | 0.94 | #30 |
| 15 | Grok 4.6 | 0.93 | #29 |
Interactive version: theaggregate.ai/benchmark?slug=prophetarena-agentic · How It Works · Data refreshed daily, snapshot 2026-11-05.