Skip to content
GPT Image 2.5
GPT Image 2.5
AI models

MiniMax H3 Max

Create with MiniMax H3 Max at 480P / 768P / 1080P · 5–15 seconds. Choose a supported input mode below and describe the subject, action and camera movement.

Start creating

Reference-site example; not a verified result generated on Kitta AI.

Available settings

Text to Video / First frame / First / last frames / Reference images · Reference images: 9

480P / 768P / 1080P · 5–15 seconds

Input modes
Text to Video / First frame / First / last frames / Reference images · Reference images: 9
Available settings
480P / 768P / 1080P
Reference images
9
Duration
5–15 seconds
Generate audio
No

Create with MiniMax H3 Max

Create with MiniMax H3 Max at 480P / 768P / 1080P · 5–15 seconds. Choose a supported input mode below and describe the subject, action and camera movement.

Stage a cinematic scene
1

Stage a cinematic scene

Specify the setting, light and framing for an establishing shot or product reveal using the available video settings.

Iterate on a video idea
2

Iterate on a video idea

Try a revised prompt for a different pace or mood, then compare completed clips before choosing one.

Guide the subject with images
3

Guide the subject with images

Use up to 9 reference images to communicate the subject and look; review appearance throughout the clip.

Describe the camera movement
4

Describe the camera movement

Ask for a slow push, tracking movement or change in viewpoint to support the action in your scene.

Plan visible motion
5

Plan visible motion

Describe how a person, object or environment should move, then check interactions and continuity in playback.

Choose the output
6

Choose the output

Choose a supported resolution and aspect ratio. Inspect text, identity and fine details in the completed result.

How to use MiniMax H3 Max

  1. 1

    Describe your idea

    Specify the subject, composition, lighting and details to preserve. Use clear instructions for each change.

  2. 2

    Guide the result with images

    Text to Video / First frame / First / last frames / Reference images · Reference images: 9

  3. 3

    Choose the output

    The price updates for your selected model, settings and output count. Review it, then choose Create. Failed paid outputs follow the applicable refund rules.

Questions about MiniMax H3 Max

Which inputs are supported?

Text to Video / First frame / First / last frames / Reference images · Reference images: 9

Which output settings can I choose?

480P / 768P / 1080P · 5–15 seconds

How do I check the price?

The price updates for your selected model, settings and output count. Review it, then choose Create. Failed paid outputs follow the applicable refund rules.

Will the result match the example?

Examples illustrate creative directions. Results vary with the model, prompt and source images; inspect your own output before use.

When do I need to sign in?

Reference uploads and cloud sessions require sign-in. The workspace shows the available actions for your account. A selected visitor file is not uploaded automatically after sign-in.

Describe your idea

Specify the subject, composition, lighting and details to preserve. Use clear instructions for each change.

Ready to create with MiniMax H3 Max?

Create with MiniMax H3 Max at 480P / 768P / 1080P · 5–15 seconds. Choose a supported input mode below and describe the subject, action and camera movement.

Start creating