CaST-Bench (Full Set) - Open-Ended: leaderboard
Open-ended justification quality on the full CaST-Bench set (2026) of 2,066 six-option video questions over 1,015 videos, where the model first outputs a grounded spatio-temporal causal chain and then answers: an LLM judge rates answer correctness, logical consistency, evidence coverage and overall justification on a 0-10 scale. Higher is better.
Metric: Open-ended justification score (0-10): an LLM judge rates the free-form answer and causal chain against the ground truth for answer correctness, logical consistency, evidence coverage and overall justification, all 2,066 six-option questions over 1,015 videos; every model must first output a grounded spatio-temporal causal chain (evidence segments with per-second boxes) and then answer; native video input except the GPT-5 models, which get 1 fps frames; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 15 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 4.3 | #178 |
| 2 | GPT-5 Mini | 3.98 | #364 |
| 3 | Gemini 2.5 Pro | 3.85 | #276 |
| 4 | Gemini 2.5 Flash | 3.83 | #463 |
| 5 | MiMo-VL-7B-RL-2508 | 2.84 | #823 |
| 6 | GLM-4.1V-9B (Thinking) | 2.66 | |
| 7 | InternVL2.5-8B | 2.59 | #2370 |
| 8 | Qwen 3 VL 4B Instruct | 2.56 | #1332 |
| 9 | InternVL3.5-8B | 2.52 | #1604 |
| 10 | InternVL3.5-14B | 2.5 | #1504 |
| 11 | InternVL3.5-30B-A3B | 2.49 | #1267 |
| 12 | Qwen 3 VL 8B (Thinking) | 2.15 | #910 |
| 13 | Qwen 2.5 VL 7B Instruct | 1.94 | #1902 |
| 14 | Qwen 3 VL 8B Instruct | 1.93 | #1029 |
| 15 | LLaVA-NeXT-Video-34B | 1.39 | #4307 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=cast-bench-full-set-open-ended · How It Works · Data refreshed daily, snapshot 2026-11-21.