MiniMax

MiniMax H3

MiniMax H3 with frame control and reference-driven video

Credits
7
per unit
Duration
4-15s
variable
Resolution
2K
output
Type
text to-video
generation

About MiniMax H3

MiniMax H3 multimodal video model with first/last frame control and reference-driven generation at up to 2K.

Best For

  • +First/last frame storytelling
  • +Reference-driven clips
  • +Higher-resolution output

Not Best For

  • -Free-tier users
  • -Ultra-long clips beyond 15s

Capabilities

First Frame ReferenceLast Frame ReferenceMultiple References

Ready to generate?

Start creating with MiniMax H3 now

View Pricing

Try Other Models

K

Kling 2.5

Kuaishou

V

Veo 3.1

Google

S

Seedance 2.0

ByteDance

MiniMax H3 AI Generator - videobao