Condesion-Bench (April-September 2025 Subset): leaderboard
Condesion-Bench (2026) evaluates conditional decision-making in a compositional action space: for one trading day an LLM allocates quantities across 15 S&P 500 stocks under variable, contextual and allocation conditions, and its single zero-shot action is checked against the conditions. This board is the decision satisfaction rate in percent on the trading days from April to September 2025, chosen to postdate recent models' knowledge cutoffs; higher is better.
Metric: Decision Satisfaction Rate (%, from a 0-1 share times 100) on the 126 Condesion-Bench instances from April to September 2025, sampled to postdate the knowledge cutoffs of recent models (zero-shot, temperature 0, JSON output); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Grok 4 | 89.68 | #321 | |
| 2 | Gemini 2.5 Pro | Medium | 88.8 | #280 |
| 3 | Grok 4 Fast | Reasoning | 84 | #446 |
| 4 | Claude Sonnet 4.5 | Thinking | 59.68 | #267 |
| 5 | Claude Haiku 4.5 | Thinking | 56.8 | #587 |
| 6 | Gemini 2.5 Flash | Medium | 43.2 | #466 |
| 7 | Nova Pro | 36.8 | #1569 | |
| 8 | Condesion-Bench Nova-Flash (checkpoint unspecified) | 26.4 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=condesion-bench-april-september-2025-subset · How It Works · Data refreshed daily, snapshot 2026-11-22.