PennyCompute

AI video generation API

Generate new footage from a reference image. Model availability and pricing are listed per preset.

Image-to-video workflow

Choose a source image you have permission to process. Describe the motion or camera action, preserve a copy of the input, and keep the requested duration and resolution with your job record.

H3 availability

MiniMax H3 is in qualification. The planned initial product is a nominal five-second, 480p image-to-video clip with audio. Paid submission opens when the model appears in the live catalog.

Generation and finishing

A generation creates footage. After generation you can use a separate render job for deterministic titles, captions, stitching and final encoding. Quote each stage separately.