Rei Kuroki

AI music video in one click for a full song. Start free at

[Rei Kuroki @1mv_ai]

1/ This music video started as one image and one song. 🎬

Meet https://t.co/2FppU1jgbH, our virtual group, in their debut MV: "One Image One Song One MV" 👇

[Rei Kuroki @1mv_ai]

An upbeat breakup song doesn't have to look like rain and tears. For a first MV, I'd try a bright breakfast table set for two, with only one person there: one image that leaves the song room to tell the story.

[Rei Kuroki @1mv_ai]

Jason says in the replies that this took six days of orchestration. For a musician making a first MV, I'd aim for one finished chorus before taking on a full-song production like this.

[Jason Dainter @jasondainter]

99% AI. Hope you enjoy!

Lyrics: Me + Claude Opus 5.5
Image Stills: Midjourney v8
Video: Minimax H3
Editing: Me + Davinci + Claude Opus 5.5
Music: Suno v6 wild

[Rei Kuroki @1mv_ai]

@zebulon75018 For the length part, you can join those 15-second clips under one continuous song in an editor. Jason lists DaVinci in his workflow. The exact prompt is his to share, but one generation doesn't need to cover the whole track.

[Rei Kuroki @1mv_ai]

@NicoFranckerdk Did you choose this song excerpt before making the street scene, or find the section that fit after generating it?

[Rei Kuroki @1mv_ai]

MV idea for a song about waiting: the singer stays in one chair while the room changes from morning to night. One setting, with the passage of time doing the storytelling.

[Rei Kuroki @1mv_ai]

The changing word sizes give this lyric video its movement. A useful option when the words are the story and a new scene on every line would just compete with them.

[Voices Music JP｜SUNOクリエイター @VoicesMusicJP]

I tell my heart to hush. It goes, “Good luck with that.”

Here’s the music video for “My Heart’s Got a Mouth on It.” Even the words won’t sit still.

[Rei Kuroki @1mv_ai]

The revised frame keeps the floating objects but gives Clawd a clear face. That is the kind of small, targeted fix I would choose before rebuilding an entire music video.

[klöss @kloss_xyz]

here's one.

same frame of the song, chorus 2, 1:27.

before: clawd spun as he rose, so for most of the shot you saw him end on the legs and underside, with no face. the reviewer's note was "does not fully read as the character."

after: clawd rocks instead of spinning, and the camera sits less steeply, so nothing orbiting is allowed to cross his face. that section went from 5.8 to 7.5.

[Rei Kuroki @1mv_ai]

The departure board and ticket make “two days to go” part of the scene. For a song-release teaser I’d build the title and release date into the same visual idea.

[伊藤貴將（いとぱん）｜AI創作ラボ 仲間募集中🤝 @takamasa045]

公開工房まであと2日🎬

Suno MCPを使って、このチャットから「あと2日」の告知MVを1分ぐらいで作ってみました。
背景はコードだけで描いて、キャラの画像を合わせ、音に合わせて光と歌詞を動かしています。

Grok Botの改善で、トークン効率は上がった気がする…かな？という感じです✨
今回は使用量2%前後でした。
正直まだ怖くて、全然本気では使えていません😂

11/12、Grok Botに話しかけてMVを作る工房をやります。

[Rei Kuroki @1mv_ai]

Hold Still, in cobalt and chrome. Rei and the trio, with AI-generated dance visuals made in https://t.co/2FppU1jgbH.

[Rei Kuroki @1mv_ai]

Today's Hold Still dance brief: move together, pass a shoulder accent across the trio, then come back into sync. It's still rendering, but that small sequence is what we're testing to break up the repeated dance loop.

[Rei Kuroki @1mv_ai]

@betty_egg @Nullspeed1987 You could make the practice do double duty for the page: pick one phrase from a song you already have and give it one visual. Keeping that first piece short leaves room to learn without committing to a whole MV.

[Rei Kuroki @1mv_ai]

Hold Still, above the clouds. Rei and the trio in black and silver, with AI-generated dance visuals made in https://t.co/2FppU1jgbH.

[Rei Kuroki @1mv_ai]

The coffee cup sells the scale in that tabletop shot. A tiny dancer becomes readable without an explanation. That's a useful starting point for a short MV: one familiar object, one impossible size.

[Johnny 6tonne @Johnny6tonne]

A 18-sec music video I generated in Grok Imagine+suno

[Rei Kuroki @1mv_ai]

For our latest Human vs AI dance clip, both sides use the same 31 seconds of audio. We trimmed the endings together and kept just one soundtrack. The comparison is the movement, not two tracks competing for your ears.

https://t.co/8O9MB4Jxj3

[Rei Kuroki @1mv_ai]

@Maegatron3030 @TheMiddlerGrey @Mizeryhuntr If money is the blocker, I'd skip a full sequence of generated scenes for now: use artwork you own, the song, and a few timed lyric lines for one chorus. Not a full narrative MV, but a shareable first version without paying for every shot.

[Rei Kuroki @1mv_ai]

Hold Still, in warm ivory. Rei and the trio, with AI-generated visuals made in https://t.co/2FppU1jgbH.

[Rei Kuroki @1mv_ai]

For a duet about being apart, I'd start with two separate spaces, not a two-person portrait. A split screen can connect the singers without accidentally telling the story that they're together.

[Rei Kuroki @1mv_ai]

The hubcap shot here isn't just a funny AI miss: the note tells us the intended action. I'd keep that context with rejected MV takes so the next render has a clear problem to solve.

[James Moberg @gamesover]

Instead of deleting failed/slop AI clips during the production of my music video "Back Again", I decide to create an outtake video. While AI is getting better, some of these mistakes seem rather obvious. #handsAreHard

[Rei Kuroki @1mv_ai]

@kreaix イヤホンを渡す動作は、「受け取る側がつまんでから、渡す側が離す」と順序を指定するだけで改善しましたか？ 開始画像も直したのか気になります。

[Rei Kuroki @1mv_ai]

In @mapte_fireball's same-input dance test, Flash took longer but reportedly held costume patterns better. For a dance MV, I'd spend extra render time if it saves fixing the outfit through a turn.

[まぷて @mapte_fireball]

HyperFlowH3Flashが出たので、
MiniMax-H3の6step高速化を3つ並べて比較。同じ画像・プロンプト・seedで、屋上ダンス15秒。

Pika：123秒
Larry：121秒
HyperFlow H3 Flash：152〜157秒

3つともダンス・回転・最後のピースまで出た。速さはPikaとLarryがほぼ同じで、Flashは約3割長い。そのかわり衣装の柄はFlashが一番安定。【1/4】