MiniMax H3 Prompt Examples: Reusable Text-to-Video Starters
A small set of MiniMax H3 prompt patterns with parameters you can reuse online. Not an auto-prompter product — practical starters only.
MiniMax H3 Prompt Examples
This page is a small experimental set of reusable MiniMax H3 starters. Search demand for “提示词 / auto prompter” on this site is still tiny, so we are not shipping a complex auto-prompter product yet.
Each example includes:
- Prompt text
- Suggested duration and aspect ratio
- A one-click path back to the online workspace handoff
Third-party note: minimaxh3.net is an independent MiniMax H3 topic hub. It is not the official MiniMax website. Official model details: MiniMaxAI/MiniMax-H3.
How to reuse an example
- Copy the prompt (or open the homepage and paste it).
- Keep model = MiniMax H3.
- Match duration / aspect when possible.
- Continue in the studio. Drafting input can start first; final generation usually needs login and credits.
Example 1 — Cinematic product turntable
Prompt
A matte ceramic coffee mug on a dark slate table, soft morning window light, slow 180-degree turntable move, shallow depth of field, subtle steam, photoreal product commercial, stable framing, no text overlays
- Duration: 5–8s
- Aspect: 1:1 or 4:5 for social, 16:9 for web hero
- Mode: text-to-video
Example 2 — City night multi-shot mood
Prompt
Night city street in light rain, neon reflections on wet asphalt, handheld but stabilized tracking shot past parked scooters, cut to a quiet overhead shot of crossing pedestrians with umbrellas, cinematic color grade, natural motion, no logos
- Duration: 8–12s
- Aspect: 21:9 or 16:9
- Tip: keep “cut to …” explicit if you want multi-beat storytelling
Example 3 — Character-consistent walking loop (image-to-video ready)
Prompt
Keep the same person and outfit from the reference image. Natural walk toward camera on a quiet residential street at golden hour, gentle camera push-in, realistic cloth motion, consistent face identity, documentary feel, no morphing
- Duration: 5–8s
- Aspect: 9:16 for vertical
- Mode: image-to-video when you have a reference still
Example 4 — Japanese-friendly scenic still-life motion
Prompt
Quiet Japanese tea room, soft shoji light, steam rising from a ceramic bowl, slow push-in, dust motes in the air, restrained color, contemplative mood, photoreal, no on-screen text
- Duration: 5–7s
- Aspect: 16:9
- Useful when testing Japanese search → online use handoff
What these prompts do not claim
- They are not guaranteed top Arena rankings
- They are not a substitute for official Context-IR / prompt skills from MiniMax
- Studio output depends on current connected model settings, credits, and moderation
Official model docs mention dedicated prompting guidance and hosted Context-IR pieces. Prefer primary sources when you need exact upstream behavior.
Related guides
Next
If an example works, save the exact prompt + duration + aspect that succeeded. That beats collecting dozens of untested marketing templates.