MiniMax H3 Max text-to-video, MiniMax's fast-generation variant of H3, turns a written brief into a 5 to 15 second clip with generated audio. Choose 480p or 768p and one of six aspect ratios from 21:9 to 9:16.
Filtering differs by endpoint: Image to video — Softened, Video extension — Softened, Text to video — Filtered.
Workbench
These examples were written for this endpoint. Replace the prompt and sample media URLs with your own accessible inputs, then check the current quote and model availability before submitting.
- Output5s
night-train-beckon Materials are the test here: green velvet, polished brass and burl walnut in one compartment, at an extremely shallow depth of field, while platform lights sweep bars across her face and the satin over and over. She holds the lens the whole way and leans in until her shoulder fills the frame. The brief also keeps her wardrobe unchanged throughout.
- 768p
- 16:9
- 5s
- Output8s
snow-train-window-hands `aspect_ratio` is `21:9`, the widest of the six frames this endpoint offers, and the shot is built for it: a train crossing the full width while two figures hold one point of contact against the glass. Written from nothing at `duration_seconds` `8`, long enough for the train to start moving and the hands to separate. Choose the frame before you write — this composition does not survive being cropped squarer.
- 768p
- 21:9
- 8s
- Output5s
silver-corridor The 21:9 result from one paragraph - fabric movement, rim light and the final look to camera are all written, none supplied as an image.
- 768p
- 21:9
- 5s
- Output5s
roller-rink A full period look asked for by name - halation, thick grain, gate weave - plus a two-part action in five seconds: a lap, then a reverse coast into camera.
- 768p
- 16:9
- 5s
This example starts at $0.05 per output second. The Playground updates the estimate when you change the settings.