実際の動画で学ぶ MiniMax H3 プロンプト集

Maru999 · @maru_999_

A Kairos H3 anime music video follows a hero and eight cafe companions through a journey of meetings and farewells.

そぴちゃんねる · @souhiro_meem_ch

An SRPG editor attaches H3-generated MP4 clips to skills as special-move sequences, alongside separately generated Suno music.

ヨンサン｜不登校系AI講師　43歳の今 · @yonsan434343

H3 animates the opening clock gears in a mixed trailer whose gameplay is captured directly from an AI-assisted roguelike.

AI Bard Guild · @IsekaiBardGuild

A ComfyUI node extracts viewpoints from an H3 orbit video, estimates poses and depth, then corrects coordinates for Qwen pose control.

Gary · @garylau_ai

A widely viewed H3 music parody depicts technology figures singing Uptown Funk, made on two DGX Sparks according to its creator.

87Boss AI 工作流實驗室 · @87BossAI

An H3 workflow combines a Selflift second pass with a 2× VAE to produce standard and enlarged versions in one pipeline.

maharshi · @maharshii

A creator reports using H3 Max to make fifteen-second explainer videos for over 300 results in a mathematical catalogue.

gokaygokay · @gokayfem

H3 Max fifteen-second repository explainers sit in a navigable 3D map of 1,000 GitHub projects clustered by shared topics.

Graalito · @Graalitoo

An H3 Max video-reference experiment achieves synchronized dual viewpoints once, while its author reports many failed attempts.

aicreataro · @aicreataro

H3 character footage drives a Blender and Python composite with audio-reactive depth, motion particles and frozen-time camera orbits.

Kamimoto(かみもと) · @sep_is_heim

A four-step H3 pipeline enlarges latents after step three and decodes with X2 VAE; its author reports 116.6 seconds for a 14-second clip.

AIossansan · @toraaiuser2

A creator passes images directly from Codex to ComfyUI for H3 animation, removing repeated manual prompt copy and paste.

sorastg524 · @sorastg524

A monochrome wireframe-world film combines H3 Max video output with CapCut AI HD enhancement.

中崎工房 | AIで1時間の仕事を5分で終わらせる人 · @nakazakifam

A fictional 2 a.m. television broadcast dated October 11, 1997 combines H3 video with Opus and ElevenLabs.

さわ|AI画像生成で月3桁継続中 · @AI_swwww

An H3 and Seedance comparison reports early dialogue drift; added acting instructions improve a bread catch while time-stop remains absent.

CANINGUM（カニンガム） · @CANINGUM

An H3 Max fan birthday short stages a small gift from a leader to Roton using Tsukishoku Kitan character materials and ElevenLabs.

めたまむ☺︎ 絵本アプリ「えほんのこや」 · @metamom_mama

A local app promo turns hand-drawn art into clay-style composites, animates them with H3 and inserts phone-screen footage in post.

元荒川ちゃん❤ · @mtArakawa

An H3 group-flight test adds individual character instructions after unspecified actions made everyone move identically.

iGavroche · @AiGavroche

A Strix Halo developer reports Veda sparse attention reducing an eleven-second H3 run from 691 to 366 seconds in ComfyUI.

wasabi_vaio · @WasabiVaio

An anime-background pipeline creates Anima stills, animates them with H3 and makes loops using FramePackVideo2LoopVideo.

neco - AI Anime · @neco1751662

A timed paper-storybook prompt takes a baby dragon from failed fire breathing to rainbow bubbles and a tiny heart-shaped flame.

⛏️🌱Don Kackman 🌱⛏️ · @Dkackman

An all-local demonstration uses Qwen 3.8 27B in OpenCode to prompt MiniMax H3.

ぐうたらするめ · @guutara_surume

A comic music video uses local H3 animation, Suno music, ChatGPT images, Real-ESRGAN and manually adjusted Whisper subtitle timing.

庭にいるルチン · @ruchiru312

An EDM creator uses H3 for a low-budget music video and reports unsatisfactory dancing and uncertainty about choreography instructions.

minmin みんみん｜AI映像作家 · @minmin_4410

An anime-ending MV adapts a credited character-controlled typography template, combining the creator’s lyrics with Suno music and H3 video.

Hong.Aigc · @newdvn2025

An open-source AutoDL skill connects H3 shot planning, six- or nine-image references, first/last frames and image-plus-audio references.

DeCat · @DeCat2025

A three-part fifteen-second horror prompt drains a dusk beach, reveals still figures on the seabed and ends with a synchronized turn.

FoldingEnd A.I. · @EndFolding79421

An H3 study compares an experimental cyberpunk LoRA on and off at twenty and eight steps, then adds a thirty-step nightclub render.

BUSET · @bseti

A continuous dolly-back prompt times a match-lighting action, a backward toss and a final reveal of giant fire lettering.

superfang119 · @superfang119

A five-beat Chinese prompt stages a solo Y2K idol teaser with close-ups, synchronized winks, sung lines and a liquid-chrome title.