Causal-Plan-Bench - Counterfactual Outcome: leaderboard
An open-ended task of Causal-Plan-Bench (2026): models state the direct consequence if a key condition does not hold, over 100 held-out video items; a GPT-5.6-sol rubric judge scores each answer against a human-verified reference on a 0-100 scale. Items come from egocentric and robot video and every model sees the same instruction and evidence; higher is better.
Metric: Open-ended counterfactual outcome (the consequence if a key condition fails) task, rubric score (0-100): a blind GPT-5.6-sol judge scores each open-ended answer against a human-verified reference from 0 to 1, times 100; 100 held-out expert-reviewed items per task drawn from egocentric and robot video (EPIC-KITCHENS, Ego4D, HoloAssist and others), the same instruction, visual evidence and answer format for every model, temperature 0.2, mean of three regenerated runs; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 46.3 | #6 |
| 2 | GPT-5.6 Sol | 37.9 | #30 |
| 3 | Gemini 3.1 Pro (Preview) | 36.3 | #94 |
| 4 | Seed 2.1 Pro | 35.7 | #135 |
| 5 | Qwen 3 VL 8B Instruct | 33.4 | #981 |
| 6 | GPT-4o | 32.6 | #741 |
| 7 | Kimi K2.5 | 31.5 | #256 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=causal-plan-bench-counterfactual-outcome · How It Works · Data refreshed daily, snapshot 2026-11-13.