T2VA · 8s · 16:9
Studio Product Reveal — T2VA
Text-led Studio Product Reveal built to rotate from texture detail to a clean pack shot without any supplied visual reference.
Editorial prompt library
MiniMax H3 Prompts for Text, Keyframe, and Reference Video
Original MiniMax H3 prompts for text-led, keyframe-led, and reference-video production.
100 TapVid-authored editorial templates following the official H3 structure, available in seven complete languages. These templates are not independently benchmarked outputs.
Explore promptsPrompt builder
Required references: —
Duration: 8s
Prompt preview
integrated_multimodal_description: From text alone, stage {{SUBJECT}} in {{SETTING}}. Start with a purposeful establishing image, let {{KEY_ACTION}} drive the camera transition, and resolve on {{OUTCOME}}. Build visual continuity through lighting, lens distance, and the scenario's material logic rather than a reference asset. This treatment presents Studio Product Reveal: the hero product in a sculptural studio set. The sequence should rotate from texture detail to a clean pack shot to create a clear purchase-ready reveal.
overall_soundscape: Build {{SOUND_PALETTE}} around the physical space and each meaningful movement; leave deliberate breath between beats.
non_diegetic_music: Use {{MUSIC_STYLE}} as a restrained progression that supports the reveal without masking the natural soundscape.Featured concepts
Start with a scenario, then set your approved production variables.
I2VA · 7s · 1:1
Feature Macro Story — I2VA
First-frame-led Feature Macro Story that extends one supplied picture into a memorable proof of the feature.
FL2VA · 10s · 16:9
Dashboard Feature Flow — FL2VA
Two-keyframe Dashboard Feature Flow that writes the transformation between a supplied opening and ending image.
L2VA · 9s · 9:16
Mobile App Onboarding — L2VA
End-frame-led Mobile App Onboarding that designs a motivated approach into the supplied final picture.
Ref2VA · 11s · 16:9
Data Story Timeline — Ref2VA
Reference-directed Data Story Timeline that assigns distinct roles to supplied subject, image, video, and audio assets.
Mode guidance
Official MiniMax H3 guidance: 4–15 seconds, native stereo audio, and Ref2VA supports up to 9 images, 3 video clips, 3 audio clips, and 12 files total.
T2VA
Author the scene from text only.
I2VA
Continue from one supplied first frame.
FL2VA
Bridge a supplied opening and ending frame.
L2VA
Design a motivated approach into one supplied final frame.
Ref2VA
Assign distinct roles to supplied subject, picture, video, and audio.
For agents
Load a static locale bundle and retain all machine identifiers and reference tokens.
Editorial guidance, not a guarantee of model output.
Official H3 format
These editorial prompts retain the documented MiniMax H3 prompt structure.
Official MiniMax H3 prompt-writing sourceTroubleshooting
- Check that every required reference token is attached.
- Keep factual claims as approved production variables.
Frequently asked questions
Are these prompts benchmarked?
No. They are original TapVid editorial prompts and are labelled editorial-unbenchmarked.
Can I change variables?
Yes. Keep approved facts, legal wording, and supplied asset identities intact.
Need an explanatory video instead?
TapVid is an Explainer Video Engine for accurate asset and information handling; it does not claim native MiniMax H3 generation.
Open TapVid