Native 2K multimodal video generation

MINIMAX HAILUO H3

Start with a scene or an image. Describe the action, camera movement, and style, then create your video in PhotoGPT.

By MiniMax

One context. Every modality. Precise control throughout. MiniMax Hailuo H3 creates coherent 2K video from text, images, and video references with controlled editing, typography, and interfaces.

Create with MiniMax Hailuo H3

Reference to Video

Text, Images, and Video in One Context

Pass up to 9 images or 3 video clips together in a single generation. MiniMax Hailuo H3 reads identity, performance, camera movement, composition, and editing rhythm from whatever you give it, then carries all of it through to one coherent result.

Precise Editing

Change One Thing, Keep Everything Else

Swap a product, rewrite signage, replace a line of dialogue, relight a scene from day to night, or add and remove objects. Localized edits land where you asked for them while the rest of the shot stays stable, so you can keep iterating on footage you already like.

Text & Interfaces

Typography, UI, and Graphics That Hold Together

MiniMax Hailuo H3 renders legible type, credits, subtitles, and brand marks, and it animates real interface work: product landing pages, game menus and HUDs, motion posters, and dynamic typography. Long, detailed prompts keep a full shot list in one request.

Character Consistency

Keep One Character Consistent

MiniMax Hailuo H3 accepts up to nine reference images and three reference video clips in a single generation, and holds the character's face, styling, motion, and voice across every shot. Pair it with a saved character from OpenArt Characters to keep the same person through a full sequence.

Prompting MiniMax H3

See how a few precise choices shape one H3 shot

Setting

Cliff city carved in stone above a waterfall-fed valley.

Camera

One continuous lock-on shot dives and threads the city.

Chase

Last-inch obstacle dodges keep the chase moving fast.

Reveal

Rainbow mist fills the open valley after the final bend.

MiniMax H3 Prompting Guide
MINIMAX HAILUO H3

One Context. Every Modality. Precise Control Throughout.

Bring text, images, and video clips into one generation, then iterate with precise edits while MiniMax Hailuo H3 preserves identity, composition, motion, and typography.

Up to 9 reference images or 3 reference videosFirst and last frame controlLocalized edits with stable surrounding footageLegible typography, subtitles, UI, and graphics
Create with MiniMax Hailuo H3

Subscription Plans

Choose the plan that works for you

Pricing plans

Starter

For getting started

$16.66$19.99

per month · billed $99.99 every 6 months

5,000credits

per month · ≈ 500 AI images

Up to 3 model trainings per month

  • 5,000 shared credits per month
  • Use credits for up to 92 Nano Banana Pro images or 250 GPT Image 2.5 medium images
  • Create up to 8 Seedance 2.5, 11 Kling 3.0 Pro, or 14 Grok Imagine clips (5 sec)
  • Generate up to 500 standard AI images and AI photos every month
  • Refine up to 500 photos with AI editing every month
  • Starter credit pool for lighter AI video, editing, and upscaling projects
Most Popular

Hobby

For regular creators

$24.99$33.3225% off sale

per month · billed $149.99 every 6 months

9,000credits

per month · ≈ 900 AI images

Up to 6 model trainings per month

  • Everything in Starter, plus 4,000 additional credits per month
  • Added credits alone cover 2 model trainings, 74 Nano Banana Pro images, or 200 GPT Image 2.5 medium images
  • Or 6 extra Seedance 2.5, 9 extra Kling 3.0 Pro, or 11 extra Grok Imagine clips (5 sec)
  • Generate up to 900 standard AI images and AI photos every month
  • Refine up to 900 photos with AI editing every month
  • More credits for regular AI video, PhotoGPT Flow, and upscaling workflows
Most Powerful

Pro

For power users

$33.33$44.4425% off sale

per month · billed $199.99 every 6 months

16,000credits

per month · ≈ 1600 AI images

Up to 10 model trainings per month

Included with every plan

Credits work across AI images, AI videos, model training, editing, upscaling, and PhotoGPT Flow.

Frequently asked questions

Start Exploring

Join our Discord community of 1000s of active users, get insights on latest prompt ideas and explore community generations

Generate VideosJoin Discord
Google Veo 3.1Seedance 2