All work fal.ai · Spec Work · 2026 30 / 31

fal — H3 Max Text to Video

A 22-second concept explainer for MiniMax H3 Max on fal: type the prompt, set the parameters, run, and direct the scene.

Unofficial concept. Not affiliated with or endorsed by fal.ai.

Companion piece to the Seedance explainer, this time for the MiniMax H3 Max text-to-video endpoint. Light theme, purple accent: a warrior-monk prompt is typed and expanded, duration, resolution and aspect ratio are set, Run fires, and the result panel fills with the generated shot before the tagline: direct the scene. The shot in the result panel is a real H3 Max generation on fal from that prompt, not a mock-up.

Same rig, second model

Reuses the After Effects playground rig from the Seedance film — world-space UI, one camera, typed prompt — restyled to the light theme so the pair reads as a series: two models, one product, one way of showing it.

Storyboard

Storyboard frames from the H3 Max explainer

More work

All work (31) ↗