Text, image and references
Use text-to-video, image-to-video, or reference-to-video according to the material available for the project.
A cost-efficient MiniMax H3 Max studio for text-to-video, image-to-video and reference-led creation with native audio.
High-value video generation
H3 is designed for frequent iteration and cost-controlled production. Generate clips up to 15 seconds, choose the faster Turbo route for routine work, and move to Seedance 2.5 when a project calls for greater visual refinement or more advanced reference control.
Review, download and reuse videos generated with H3.
MiniMax H3 Max capabilities
Begin with a written prompt, animate a first frame, or guide the result with multiple reference assets. H3 keeps these workflows in one consistent studio.
Use text-to-video, image-to-video, or reference-to-video according to the material available for the project.
Select output duration, resolution and aspect ratio for social posts, advertising concepts, storyboards and production drafts.
H3 can produce synchronized audio with the video, reducing the need to assemble a separate sound track for early creative iterations.
Describe the scene, action, camera language and sound. H3 generates the complete clip from the prompt.
Use a first frame to establish composition and appearance; an optional end frame can guide the final visual state.
Combine images, video and audio references, then mention uploaded assets with @ references to guide subjects, motion and sound.
Both routes use the H3 family. The choice affects generation speed and credit consumption; reference-to-video uses Max.
Use the standard H3 Max route for image, text and reference-led generation, especially when the project depends on multiple reference assets.
Use Turbo for faster and lower-cost text-to-video or image-to-video iteration. The optimized route maintains output quality while improving throughput.
Key information about MiniMax H3 Max inputs, output settings, modes and credit use.