Cinematic AI video from text or a first frame, at 768P or 1080P.
Sample outputs. Open in Studio to generate your own.
Turn a written scene description into a short, cinematic-quality video clip.
Animate a still by uploading it as the start frame to switch into image-to-video.
Guide a clip with an optional first and last frame so it begins and ends where you want.
Hailuo 02 is MiniMax's cinematic AI video model, built to turn a written prompt into motion. Describe a scene and the model renders it as a short video clip, with a positioning aimed at cinematic-quality output.
The model works two ways. Start from text alone for pure text-to-video, or add a start image to anchor the first frame and switch into image-to-video. You can also supply an optional end image to guide the last frame, giving you control over where the shot begins and where it lands.
Pick your output tier to match the job. The Standard tier runs at 768P and can stretch to 10 seconds, while the Pro tier delivers 1080P for sharper 6-second clips. On HexGen, Hailuo 02 runs in your browser with no setup, and you pay per video by the resolution and duration you choose.
Copy, tweak, and run. Good prompts get you most of the way there.
A lone surfer paddling out at dawn, slow camera glide over glassy waves, soft golden light
A vintage train pulling into a misty mountain station, steam drifting across the platform
Close-up of rain hitting a neon-lit city window at night, cinematic depth of field
Pay only for what you render. 1 USD = 1,000 HGcoins. HGcoins never expire and failed runs refund automatically.
How Hailuo 02 compares with other video models on HexGen.
| Model | Quality | Speed | Cost | Choose it when |
|---|---|---|---|---|
Hailuo 02 This MiniMax | Best | Fast | Mid cost | You want cinematic video from text or a first frame, with optional start and end frame control |
Kuaishou | Best | Fast | Mid cost | You want fine-grained control modes and strong prompt adherence |
ByteDance | Best | Fast | Higher cost | You want the most cinematic motion at the highest tier |
It is MiniMax's cinematic AI video model. It generates short video clips from a text prompt, and can also work from an image when you supply a start frame.