AI video answer
Which MiniMax H3 prompt mode should I use?
Use T2VA for text only, I2VA for an exact first frame, FL2VA to connect exact first and last frames, L2VA to land on an exact final frame, and Ref2VA for full multimodal references.
Choose text or a keyframe mode
Choose T2VA when words define the full shot. Choose I2VA when Picture 1 is the exact opening frame and the action develops forward from it.
Choose FL2VA when both endpoints must match. Choose L2VA when the model must invent a plausible opening and converge on the supplied last frame.
Use full-reference mode for relationships
Choose Ref2VA when images, video, or audio control subjects, motion, camera language, editing, continuation, voice, music, or other reusable traits.
The official reference format includes subject definitions, a task summary, retention analysis, detailed description, soundscape, and music.
Do not mix the schemas
Base and keyframe modes use the three core fields. Full-reference mode uses six ordered sections and stable labels such as Picture 1, Video 1, and Audio 1.
Pick the mode before writing. This prevents a reference relationship from being lost inside a text-only format.
Related guide
Get the full context.
MiniMax H3 prompt guide and templates
Write MiniMax H3 prompts for 4–15 second video up to 2K with native stereo sound, keyframe control, and image, video, or audio references. Copy three templates.
Read the minimax h3 prompt guide guideCurrent prices
Compare each video route.
| Model | Compared rate | Cre8vid price | 5 sec | 10 sec | Savings |
|---|---|---|---|---|---|
| MiniMax H3MiniMax | $0.065 | $0.029 | $0.145 | $0.29 | 55% less |
| Wan 3.0Alibaba | $0.050 | $0.025 | $0.125 | $0.25 | 50% less |
| Kling 3.0Kuaishou | $0.084 | $0.039 | $0.195 | $0.39 | 54% less |
| Seedance 2.5ByteDance | $0.144 | $0.069 | $0.35 | $0.69 | 52% less |
| Veo 3.1 Fast + audioGoogle | $0.100 | $0.049 | $0.24 | $0.49 | 51% less |
Source notes