Ima Studio · @ImaStudio_ai
Midnight Seoul Fashion Film
Eight fashion looks, locations and typography beats compressed into a timed 15-second transformation film.
經查核的提示詞與原始出處
對照創作者公開的提示詞與完成的 MiniMax H3 影片,了解場景安排、動作、運鏡、連貫性與聲音如何搭配。
每個案例都附有原始出處連結。你可以直接複製公開的提示詞,也能保留其結構,改寫成自己的場景。
706 個經查核案例
依運鏡技巧、分鏡、動畫、參考控制、音樂、對白或動態圖像瀏覽案例。已公開提示詞與僅介紹工作流程的案例皆有清楚標示。
找到 706 個案例 · 顯示第 1–30 筆
Ima Studio · @ImaStudio_ai
Eight fashion looks, locations and typography beats compressed into a timed 15-second transformation film.
Kōda · @aimikoda
A ten-shot character reveal whose contour lines, graphic accents and music adapt to the supplied design.
Pierrick Chevallier · @CharaspowerAI
A dense single-take FPV path uses macro texture, parallax, occlusion and chained rack focus around one subject.
Saul Goodman · @Goodmanprotocol
A four-shot, 15-second night chase specifies vehicle physics, wet-road lighting, camera changes and Doppler audio.
Glenn Williams · @GlennHasABeard
A minimal causal instruction turns each popped bubble into one slowly rising firefly across a locked shot.
Glenn Williams · @GlennHasABeard
One directional constraint keeps a sweater shrinking while the same yarn grows into a scarf in a locked composition.
Glenn Williams · @GlennHasABeard
A compact rate constraint makes marbles leave a jar in single file, roughly once per second, without pairs.
Glenn Williams · @GlennHasABeard
A distance-over-time instruction controls slow motion by making one feather cross one window in twelve seconds.
852話(hakoniwa) · @8co28
A single 3×3 reference sheet becomes a 15-second animation that advances panel by panel without split screens.
さんかくてん · @sankakuten91256
One square sheet of sixteen illustrated beats drives a compact anime sequence while retaining much of the source composition.
ぽんず | AI映像 · @ponzponz15
A single local-event flyer is reinterpreted as a 15-second social video with layered motion-graphics treatment.
ぽんず | AI映像 · @ponzponz15
The familiar folktale becomes a graphic 12-second sequence that compresses characters and story beats into bold motion design.
AIスタジオワンルーム(AIアニメ、動画、漫画) · @studio_oneroom
A polished 15-second motion-graphics study demonstrates an animation workflow made directly from text without post-editing.
中崎工房 | AIで1時間の仕事を5分で終わらせる人 · @nakazakifam
A 1080p text-to-video fashion campaign combines rapid outfit changes, editorial layouts and Japanese narration.
Shim · @os_shim
A longer-form teaser assembles MiniMax H3 cinematic shots with music into a compact film-trailer presentation.
ウミノ🤖AI×マーケティングのプロ · @umino_chibi
A creator turns a recurring cute-grotesque parakeet character series into a set of animated messaging stickers.
Victor M · @victormustar
Character clips for eight moves are keyed from a magenta background, trimmed into pose extremes and packed as a playable sprite atlas.
Maurice Bourdon · @MauriceBourdon
Multiple locally generated text-to-video segments are context-connected into a continuous two-minute visual loop.
ComfyUI · @ComfyUI
A basic motion blockout and reference images guide a polished vertical shot, with an optional second pass upscaling to 2K.
gokaygokay · @gokayfem
An interactive world pipeline turns each autonomous decision into video, producing 264 clips for a city carried by a whale.
AIライフハック · @ai_lifehack55
Three heavy paint forms leave a billboard, race through a city, then restore the original printed design in six timed shots.
Kōda · @aimikoda
Thirty-plus hard cuts carry one character action across detail, portrait and full-body angles while preserving identity and momentum.
Mr Das · @MrDasCreates
Four structured JSON prompts build a coherent wildlife reel of invented animals sharing one photoreal pink salt-lagoon habitat.
不満福 · @fumanfuku
A reference walking video transfers runway motion to a consistent stylized character in a 15-second vertical R2V result.
知识猫AI实验室 · @GeekCatX
A video reference is converted into seamless matte-gray clay models while preserving every subject, action, beat and camera move.
Kōda · @aimikoda
One luminous ribbon and an uninterrupted orbit reveal a character from feet to face before a colorful full-body hero payoff.
AI Bard Guild · @IsekaiBardGuild
A standing character retrieves an electric guitar, sings supplied Japanese audio and performs with synchronized mouth movement.
Loriel.AI · @ou_zhen599
A single-room 15-second scene escalates a potion mistake through precise acting, viscous fluid physics, camera limits and dialogue.
Umesh · @umesh_ai
A 15-second neo-noir opener embeds credits into rain, brass, glass and street reflections before revealing its main title.
Pierrick Chevallier | IA · @CharaspowerAI
Successive energy cuts forge a steel BLADE title from flying titanium fragments before one final slash destroys it.
使用合集
將每個案例視為可重複運用的結構。保留控制動作與連貫性的部分,再替換原始場景中的創意細節。
瀏覽分類,或依主體、運鏡技巧、創作者與模型搜尋,選擇動作及結構符合預期影片的案例。
若有公開文字,請先確認是完整提示詞還是片段再複製。沒有提示詞的案例,可參考原始出處與改寫建議。
換成自己的主體與場景,提供所需的參考素材,並在生成前檢查所有限制條件。
複製經查核的案例、調整創意細節,使用 MiniMax H3 生成自己的影片。