mcps
S K D

Recipe

Make a lyric video from your song

Write the song in Suno with bracket-tag sections, generate one visual per section in Kling, and let the song's own structure drive the shot list.

for Anyone who wants a song, a jingle, or a voice — no instruments, no studio

Hand-checked recipe · curated by Ilya Gindin

~60 minutes

to run

intermediate

difficulty

3

connectors

What you'll need

Suno Generates the song — and its bracket-tag sections become the storyboard.
Required
Kling Renders one looping visual per song section, consistent in style.
Required
Descript Optional — assembles sections and lays the lyrics on screen as captions.
Optional

This recipe runs in 4 steps

  1. 1

    Write the song with explicit sections

    Say to your AI

    Write an upbeat indie-pop song with these sections: [Verse] city morning, [Chorus] big sing-along hook, [Verse] night drive, [Chorus] louder, [Outro] fade.

    You'll get: A Suno track whose sections you named — that's your storyboard.

  2. 2

    One visual per section

    Say to your AI

    Generate a 5-second loop for the verse: city morning, warm film look, slow drift. Same style for every clip: warm film, gentle motion.

    You'll get: A Kling clip per section, consistent in look.

  3. 3

    Assemble to the song's structure

    Say to your AI

    Lay the clips over the track in section order — verse visual for the verse, chorus visual for both choruses.

    You'll get: A rough cut where the visuals switch on the song's own seams.

  4. 4

    Lyrics on screen

    Say to your AI

    Add the lyrics as large centered captions timed to the vocal, and export 1080p MP4.

    You'll get: A finished lyric video. (No Descript? Any editor works — clips and track are done.)

You're done when

A lyric video where the visuals move on the song's own structure — verse to chorus — because the bracket tags that shaped the track also shaped the shot list.

The chain is Suno → Kling → Descript, about an hour: write the song with named sections, render one consistent visual per section, assemble on the song’s own seams, put the lyrics on screen. The trick that makes it work: the bracket tags you give Suno ([Verse], [Chorus]) are simultaneously the arrangement AND the storyboard — one prompt decision drives both tools.

That’s seam knowledge no README carries: Suno’s docs treat tags as music structure, Kling’s docs don’t know your song has sections. Chained, they turn structure into a shot list for free.

Checked 2026-07-12: Suno and Kling community connectors reachable (each needs its own token); Descript official, optional here.

If something breaks

The song came back one shapeless take
Fix: Use bracket tags: [Verse], [Chorus], [Bridge]. Suno reads them as real arrangement cues — and they double as your storyboard.
Visuals look like different films
Fix: Repeat the same style words in every Kling prompt (warm film, gentle motion). Each clip renders independently — consistency has to come from you.
The chorus visual jumps when it repeats
Fix: Ask Kling for a seamless loop with matching first and last frame, then reuse the same clip for every chorus.
I don't see Suno or Kling to turn on
Fix: Both are community connectors: add each MCP with its token — Suno through a third-party music API (sunoapi.org key), Kling with your Kling API keys.

People ask their AI

“make a lyric video for my song”“visuals for a Suno track”“song with an animated visual story”

Related

Suno Kling lo fi track with a looping visual