AI Model Shootout: Same Prompt, Every Model

The same prompts and seed through 32 image models, the same four stills through 7 video models, the same pictures through 5 image-to-3D models and the same prompts through 6 music and sound-effect models, each with its measured speed on the GPUs we test. Speeds and VRAM for every model are in the AI models database.

Image models

Prompt: “studio portrait of an elderly fisherman with a grey beard, soft window light, 85mm photo”

Image to video models

Text to video models

Image to 3D models

Music and sound-effect models