Generate 5–15 second native 2K clips from text, start and end frames, or image, video, and audio references.
Model overview
MiniMax H3 is a general-purpose multimodal video model released by MiniMax on July 31, 2026. It reads text, images, video, and audio together, then generates a short video with native stereo sound.
MiniMax H3 treats your prompt and reference media as one creative context. Use an image for a subject or visual direction, a video for motion or camera language, and audio for voice, rhythm, or sound direction.
Visuals and sound are generated in the same process instead of being assembled afterward. On this site, every H3 workflow returns a 5–15 second 2K MP4 with its generated audio track, without a separate audio toggle.
Core capabilities
MiniMax H3 combines direct shot control with mixed-media references. These five capabilities help turn a brief, keyframe, or existing asset into a more intentional short clip.
Workflow example · multimodal reference
Reference mode accepts up to 12 ordered files within the image, video, and audio limits. Refer to each asset by its numbered label, such as Image 1, Video 1, or Audio 1, so H3 knows which reference should guide the subject, movement, camera, style, voice, or timing.
Visual example · subject and style continuity
Use a clear reference image or video to help preserve a character, product, palette, or visual treatment through a generated shot. Specific prompts and clean source material give the model a stronger target.
Visual example · fast action and camera movement
A reference video can guide movement, pacing, camera paths, and editing rhythm. This is useful when the action is easier to show than to describe frame by frame.
Workflow example · first and last frame control
Upload a starting frame, an ending frame, or both to set the composition and destination. MiniMax H3 builds the movement between those frames while the prompt directs the action and camera.
Sound-on example · dialogue, ambience, and picture together
This integration outputs 2K video in whole-second durations from 5 to 15 seconds. Because sound is generated with the picture, the result can include dialogue, ambience, effects, and music in the same clip.
Three steps
Use MiniMax H3 in the browser without writing API code. Choose the control you need, add only the references that serve the shot, and check the exact quote before submission.
Use Text to Video for a prompt-only shot, Start & End Frames for a controlled transition, or Multimodal References when images, video, or audio need to guide the result.
Describe the subject, action, environment, camera path, visual style, and sound. In reference mode, mention uploaded media by order and explain the role of each asset.
Choose the duration and, in text and reference modes, the aspect ratio. Wait for uploads to finish, review the credit quote, submit the task, then save the completed video from your library.
Popular use cases
MiniMax H3 is built for short-form audiovisual work where motion, references, and sound need to follow one direction. Start with one focused deliverable per generation.
Turn product images, campaign art, and brand references into short launches, product reveals, and visual ads with controlled motion and sound.
Use character, wardrobe, and style references to develop a compact scene in live action, illustration, anime, 3D, or mixed-media aesthetics.
Animate titles, interface concepts, game menus, posters, and branded graphic systems when the design itself needs to become the subject of the video.
Prototype environments, abilities, transformations, action beats, and effect-heavy shots before committing time to a full production pipeline.
Create vertical or horizontal story beats, campaign variants, mood tests, and scroll-stopping clips for social publishing and creative review.
AI video model comparison
MiniMax H3, Seedance 2.5, and Kling 3.0 take different approaches to multimodal video creation. Compare their publicly documented focus, inputs, duration, resolution, audio, and best use cases before choosing a model.
Capabilities reflect official public documentation. Availability, duration, and output settings can vary by product, region, and API.
Model selection
Choose by the material you already have, the control you need, the output length, and the production goal. Each model emphasizes a different part of the creative process.
MiniMax H3 is a strong fit for short 2K videos guided by ordered image, video, or audio references, especially when first and last frame control matters.
Seedance 2.5 is better suited to longer scenes, high-resolution campaign assets, and production workflows that combine many references or require local editing.
Kling 3.0 is a strong option for cinematic multi-shot clips, stable characters or products, and scenes that depend on multilingual or multi-character dialogue.
Pricing and limits
MiniMax H3 generation is credit-based. A 5-second text or frame video has a 325-credit base quote before any plan multiplier. Reference-video seconds and images beyond the first five can increase the base, and the generator shows the final quote before submission.
Compare credit plansGeneration settings
The generator validates these limits before a task is submitted.
Output
Native 2K
MiniMax H3 currently uses the 2K quality setting.
Duration
5–15 seconds
Whole-second values only.
Text / reference ratio
21:9, 16:9, 4:3, 1:1, 3:4, 9:16
Image mode follows the input frame.
Reference images
Up to 9
JPG, JPEG, PNG, WEBP; up to 30 MB each on this site.
Reference videos
Up to 3
2–15 seconds each, up to 15 seconds total and 50 MB each.
Reference audio
Up to 3
MP3 or WAV, 2–15 seconds each, up to 15 seconds total and 15 MB each.
Reference files total
Up to 12
The image, video, and audio files in one request count together.
Reference requirement
Image or video required
Audio cannot be submitted as the only reference type.
Result availability
24 hours upstream
The provider download URL expires after 24 hours; completed files are copied to your library when storage succeeds.
Generate on minimaxh3.tv
Use one account to move between prompt-only generation, start and end frames, and ordered multimodal references. Each workflow shows its practical limits and exact credit quote before submission.
minimax-h3-text-to-video
Describe the subject, action, environment, camera movement, and visual style in one prompt.
minimax-h3-image-to-video
Animate a starting frame, an ending frame, or both. The output composition follows the uploaded frame.
minimax-h3-reference-to-video
Guide the result with ordered image, video, and audio references while describing how each reference should be used.
Generation workflow
Move between three MiniMax H3 workflows, review practical limits and credit costs before submission, and keep each project in one workspace.
Move between prompt-only generation, start-and-end-frame animation, and multimodal references without rebuilding the project in a separate tool.
The generator calculates the quote from duration and billable reference inputs before the task is submitted, so the credit cost is visible before you commit.
MiniMax H3 stays the default model, while supported alternatives remain available when a different brief calls for another workflow or output profile.
Selected video examples
Each refresh selects eight examples from a ten-video project library, spanning graphic animation, dialogue, character work, camera movement, and cinematic storytelling.
Reference media anchors texture and subject details while camera movement turns a simple product idea into a polished short sequence.
A restrained conversation uses alternating close-ups, stable faces, and consistent lighting to keep the scene readable.
A fast visual overview connects reference media, generated frames, motion, and sound into one multimodal creation workflow.
Defined opening and closing frames give a short scene a clear visual destination while leaving room for generated movement between them.
A quiet period scene focuses on subtle expression, eyelines, and stable close-up framing instead of large visual effects.
Bold color fields, animated type, and playful character design turn a compact visual idea into a fast graphic sequence.
Rapid character motion, changing camera distance, and a neon environment test whether a busy action beat remains easy to follow.
A glowing spirit turns an everyday laundromat into a neon wireframe world while the first-person camera keeps the transformation connected.
Questions before generation
Keep creating
Move from model research to the workflow that matches your source material, or review examples and credit plans before generating.
Guides and research
Use the practical guides for prompt structure, current browser workflows, and API task handling when you need more depth than a model overview.
Prompt examples
Build prompts around a visible first frame, one motion goal, one camera move, and a clear ending state.
Workflow guide
Follow the text, start/end frame, and multimodal reference workflows from input through download.
API guide
Review task creation, status polling, file retrieval, route IDs, and current pricing inputs.
Model alternatives
Open a related model in the same generator when your brief needs a different resolution, duration, or input workflow.
Hailuo 3 · Native 2K
Choose a workflow, review the exact credit quote, and submit your native 2K video.
