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.

137–144 of 214

ground truth

Exploring Bourbon

model outputs

Gemini 3 Flash Preview

A 0.82T 0.24

Qwen3-VL-8B-Instruct

A 0.70T 0.27

GPT-5.4

A 0.79T 0.30

Claude Sonnet 4.6

A 0.82T 0.21

LLaMA 4 Scout

A 0.64T 0.34

ground truth

Exploring Bourbon

model outputs

Gemini 3 Flash Preview

A 0.81T 0.25

Qwen3-VL-8B-Instruct

A 0.77T 0.23

GPT-5.4

A 0.90T 0.26

Claude Sonnet 4.6

A 0.87T 0.30

LLaMA 4 Scout

A 0.59T 0.29

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.91T 0.30

Qwen3-VL-8B-Instruct

A 0.54T 0.26

GPT-5.4

A 0.89T 0.33

Claude Sonnet 4.6

A 0.90T 0.29

LLaMA 4 Scout

A 0.68T 0.29

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.91T 0.25

Qwen3-VL-8B-Instruct

A 0.78T 0.29

GPT-5.4

A 0.90T 0.19

Claude Sonnet 4.6

A 0.86T 0.27

LLaMA 4 Scout

A 0.79T 0.20

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.90T 0.29

Qwen3-VL-8B-Instruct

A 0.70T 0.24

GPT-5.4

A 0.88T 0.24

Claude Sonnet 4.6

A 0.90T 0.25

LLaMA 4 Scout

A 0.67T 0.00

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.88T 0.32

Qwen3-VL-8B-Instruct

A 0.75T 0.29

GPT-5.4

A 0.91T 0.31

Claude Sonnet 4.6

A 0.85T 0.31
no output

LLaMA 4 Scout

A —T —

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.89T 0.29

Qwen3-VL-8B-Instruct

A 0.55T 0.17

GPT-5.4

A 0.71T 0.28

Claude Sonnet 4.6

A 0.89T 0.30

LLaMA 4 Scout

A 0.50T 0.22

ground truth

CSS Spinners

model outputs

Gemini 3 Flash Preview

A 0.89T 0.33

Qwen3-VL-8B-Instruct

A 0.78T 0.27

GPT-5.4

A 0.90T 0.19

Claude Sonnet 4.6

A 0.86T 0.30

LLaMA 4 Scout

A 0.85T 0.29
← Previous18 / 27
go to
Next →