Hedge-Bench - Theme Coverage: leaderboard
Metric: Themes covered (%), macro-averaged over tasks: a theme counts as covered once enough of its grounded expert moves are hit (max(1, min(n-1, 3)) of n); 102 open-ended financial-analyst research tasks derived from recorded expert analyst discussions, run in the Harbor harness with the Terminus 2 agent over a sandboxed document corpus, eight trials per task, graded by a Gemini-3.1-Pro judge against expert rubric moves with hallucination-supported moves discounted; higher is better. Source: arxiv.org. Saturation forecast: Around August 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 56.4 |
| 2 | Claude Opus 4.7 | 53.6 |
| 3 | GPT-5.5 | 48.1 |
| 4 | Gemini 3.5 Flash | 48 |
| 5 | Claude Opus 4.8 | 47.3 |
| 6 | Claude Haiku 4.5 | 32.5 |
| 7 | Gemini 3.1 Pro (Preview) | 30.5 |
| 8 | GPT-5.4 Mini | 19.8 |
Interactive version: theaggregate.ai/benchmark?slug=hedge-bench-theme-coverage · How It Works · Data refreshed daily, snapshot 2026-09-29.