animation2code benchmark
For best compatibility, please view this dashboard in a Chrome browser.

Zero-shot video (or image-frame) → code results on the test set, across commercial and open-source models.

Each output is tagged with A = appearance similarity and T = temporal similarity; higher is better for both. Click a video to inspect its code.

97104 of 214

model outputs

GPT-5.4

A 0.86T 0.29

Claude Sonnet 4.6

A 0.78T 0.33

LLaMA 4 Scout

A 0.82T 0.22

model outputs

GPT-5.4

A 0.72T 0.24

Claude Sonnet 4.6

A 0.67T 0.29

LLaMA 4 Scout

A 0.62T 0.25

model outputs

GPT-5.4

A 0.76T 0.31

Claude Sonnet 4.6

A 0.73T 0.34

LLaMA 4 Scout

A 0.65T 0.26

model outputs

GPT-5.4

A 0.80T 0.32

Claude Sonnet 4.6

A 0.85T 0.18

LLaMA 4 Scout

A 0.32T 0.31

model outputs

GPT-5.4

A 0.94T 0.27

Claude Sonnet 4.6

A 0.94T 0.29

LLaMA 4 Scout

A 0.64T 0.26

model outputs

GPT-5.4

A 0.84T 0.26

Claude Sonnet 4.6

A 0.82T 0.29

LLaMA 4 Scout

A 0.71T 0.22

model outputs

GPT-5.4

A 0.92T 0.26

Claude Sonnet 4.6

A 0.90T 0.26

LLaMA 4 Scout

A 0.68T 0.25

model outputs

GPT-5.4

A 0.85T 0.31

Claude Sonnet 4.6

A 0.85T 0.30

LLaMA 4 Scout

A 0.75T 0.30