MiniMax H3
Multimodal video generation and editing with text, image, video, and audio inputs
About MiniMax H3
MiniMax H3 is a next-generation general-purpose multimodal video model for video creation. It brings text, images, video clips, and audio into one creative context to generate video with native two-channel stereo. It supports text-to-video, image-to-video, and instruction-based video editing, with reference uploads and 5–15-second, 24 FPS, up to 1440p output.
Best for
Best use cases
Text-to-video and image-to-video
Instruction-based editing of existing videos
Advertising and brand film production
E-commerce and social content production
Features
Multimodal video generation
Understands text, images, video clips, and audio in one creative context and generates video.
Instruction-based video editing
After uploading a video, replace characters, remove objects, change backgrounds, or adjust lighting and dialogue by description.
High-spec video output
Generates 5–15-second videos at 24 FPS and up to 1440p.
Native stereo
Every video includes two-channel stereo audio.
Pros
- Understands text, image, video, and audio context in one workflow
- Provides native two-channel stereo when generating video
- Supports reference uploads and instruction-based editing
Cons
- Each request has input limits: prompts up to 7000 characters and up to 12 files in total.
Pricing Plans
Lite
- 600 credits/month
* Pricing information is for reference only. Please refer to the latest announcement on MiniMax H3's official website.