⭐ Open-source · bring your own key · no accounts

Turn a song into a
complete music video

A desktop app that writes a story from your lyrics, designs the shots, generates the keyframes, animates every scene, and cuts it to the beat — all on your machine, paid at cost with your own API keys.

Windows · macOS · Linux — free & open-source (Apache-2.0)
🎵 song transcribe story from lyrics shot list keyframes · your cast per-scene video assemble 🎬 MP4
What it does

A real music video, from the actual song

Not generic AI filler — a narrative that tracks your lyrics, locked to the singing.

📝

Story from the lyrics

The video tells the song’s own story — a coherent arc built from the real sung words, in order.

🎯

Vocal-aligned editing

Forced-aligned transcription + a frame-grid retime, so every cut locks to the singing with no drift.

🎭

Reusable cast

A photo and/or a description becomes a consistent character, kept recognizable across every scene.

🔁

Per-scene refresh

Don’t like one shot? Re-roll a single scene without touching the rest of the cut.

Preview first

Render the opening ~25% cheaply, watch it, then continue to the full song.

🎚️

Your models

Swap the LLM / image / video models in Settings — pick a cheaper engine whenever you want.

Bring your own key

No subscription. No markup.

Cloud video generation genuinely costs ~€25–30 of provider cost per finished minute. A hosted service has to mark that up. Videoboom removes the middleman: paste your own OpenRouter / Replicate keys, the app calls those providers directly, and you see exactly what each render costs. Keys are encrypted with your OS keychain and never leave your machine.

0% markup — paid at cost
0 accounts or servers
100% local — your files stay put
How it works

Drop a song. Get a video.

  1. Add your keys. In Settings, paste your OpenRouter key and Replicate token (forced-aligned lyric timing). Encrypted via your OS keychain.
  2. Drop in a song and optionally build a cast — reusable characters from a photo or a description.
  3. It writes & renders. Story from the lyrics → shot list → identity-preserving keyframes → per-scene clips → beat-cut assemble.
  4. Preview, refine, export. Watch the preview, re-roll any scene, then render the full song to an MP4 (tagged as AI-generated).

Ready when you are

Free, open-source, runs on your machine.

Builds are unsigned — on first launch, right-click → Open (macOS) or “More info → Run anyway” (Windows).