ChronoVision - Artifact Ordering: leaderboard
Artifact ordering on ChronoVision (2026), a temporal reasoning benchmark for vision-language models: in 1,000 tests the model puts artifact images from different dynasties in chronological order, scored as 50 times one plus Kendall tau (0 to 100). Higher is better.
Metric: Ordering score 50 x (1 + Kendall tau) (0-100) on the 1,000 artifact sorting tests: order images from different dynasties chronologically; zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around July 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 64.05 | #266 |
| 2 | Qwen 3 VL 235B A22B Instruct | 62.36 | #522 |
| 3 | GPT-5.2 | 62.21 | #176 |
| 4 | Qwen 3 VL 8B Instruct | 61.32 | #981 |
| 5 | GLM-4.1V-9B (Thinking) | 59.37 | #1060 (GLM-4.1V-9B) |
| 6 | Qwen 3 VL 4B Instruct | 58.23 | #1279 |
| 7 | Qwen 2.5 VL 7B Instruct | 56.78 | #1833 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=chronovision-artifact-ordering · How It Works · Data refreshed daily, snapshot 2026-11-13.