MiniMax H3: Video with Sound and Speech

MiniMax H3 makes video whose sound and speech are born together with the picture, at any length from 4 to 15 seconds. Start from text or an image, and add pictures, clips and audio as references.

Length

Any whole length from 4 to 15 seconds

Resolution

768P or 2K

Sound

Sound and speech in the same pass

Price

From 4 credits a shot

What you can do with MiniMax H3

• Text to video
• Image to video, with an optional end frame
• References: 9 images, 3 clips and 3 audio files, up to 12 files

What it is strong at

• Sound and speech in the same pass as the picture
• Priced by the second, so you pay for exactly the length you need

What to know first

• Images past the fifth, and reference clips, add to the price. Audio references are free.
• Dialogue works best when it is short.
• Hebrew speech is not born inside the shot. Hebrew narration is added separately, and a character talking on camera is lip-synced to that voice.

What MiniMax H3 costs

768P, 4 seconds

4 credits

768P, 5 seconds

5 credits

768P, 10 seconds

10 credits

768P, 15 seconds

15 credits

2K, 5 seconds

8 credits

2K, 10 seconds

16 credits

2K, 15 seconds

23 credits

Common questions

Does MiniMax H3 have speech?

Yes, sound and speech are born in the same pass as the picture. It does not speak Hebrew, so Hebrew speech is added separately.

How much does a 10 second shot cost?

10 credits at 768P and 16 credits at 2K.

How do I pay?

You buy a credit pack, and the same credits work on every engine. The price of every creation shows before it runs.

What if a generation fails?

The credits return to your balance automatically, with no need to ask.

More engines on the same balance

• Article: MiniMax H3
• Gemini Omni 1.2 Flash
• Kling 4.0
• Seedance 2.5
• Wan 3.0
• Vidu Q4
• LTX-2.5
• Grok Imagine
• GPT Image 2.6
• Nano Banana 2.1
• Midjourney V8.2
• Seedream 5.0 Pro
• FLUX 3 Image
• Ideogram 4.5
• Eleven v4
• Suno V6
• Packs and pricing
• All the engines