Tasks and inputs
MiniMax H3 is listed for Image-text-to-video, Text-to-video. It accepts text, image, video, audio and produces video, audio. Use the publisher’s input format and preprocessing for your chosen deployment.
MiniMax H3 is a multimodal generation system that produces video with synchronized stereo audio from supported text, image, video and audio references.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
MiniMax H3 is a multimodal generation system that produces video with synchronized stereo audio from supported text, image, video and audio references.
The publisher describes H3-Context-IR, H3-Base and H3-Regenerate-2K. The base stage produces 768p output; the regeneration stage produces 2K output. H3-Context-IR is a hosted preprocessing system and is not included in the open-weight release. Downloading the base weights does not reproduce every part of the hosted pipeline.
First-and-last-frame generation and multimodal reference generation use different variants and input specifications. Choose the mode before preparing the assets. Write down which reference supplies appearance, motion, sound or scene context so the result has a clear brief to satisfy.
The documented system supports clips up to fifteen seconds. Inspect the whole clip for identity drift, abrupt motion, invented details and audio that does not match the visible action. A strong first frame is not enough to assess a video model.
For local serving, use the supplied tokenizer and configuration files with the documented inference workflow. Read the MiniMax H3 Community License for your intended use. No hosted generation price is inferred from the size of the checkpoint.
Use the same clip brief, duration and target resolution when comparing with Veo 3.1. Record both accepted-output cost and the time needed to repair or regenerate a failed clip.
Publisher source 1 Deployment limits and commercial terms belong to the selected checkpoint or API route.
Compare the model on the work you need to complete.
MiniMax H3 is listed for Image-text-to-video, Text-to-video. It accepts text, image, video, audio and produces video, audio. Use the publisher’s input format and preprocessing for your chosen deployment.
Choose duration, resolution and frame rate together. Compare the cost of an accepted clip, including repeat generations.
The weights are downloadable under MiniMax H3 Community License. Check runtime support and size the complete model and working context for your expected concurrency.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
xAI
Grok Imagine Video 1.5 Lite generates video from text or an input image, with resolution-based pricing on the xAI API.
ByteDance Seed
Seedance 2.5 from ByteDance Seed: ByteDance multimodal audio-video model with up to 30-second single-pass generation, multi-round extension, reference inputs and timestamp-level editing.
Gemini Omni Flash Preview from Google: Video generation and editing model for fast, conversational text- and image-to-video workflows.
xAI
Grok Imagine Video 1.5 creates video from text, image references and supported voice inputs, with frame controls for planned shots.
Veo 3.1 Lite Preview from Google: Video model for prompt-guided generation, editing, and motion workflows.
ByteDance Seed
Seedance 2.0 from ByteDance Seed: ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.