Music → Suno Prompt

00

Recent generations

Past prompts (and their variants) saved to a local SQLite DB. Load to reuse without regenerating.

No saved prompts yet — generate one below.

01

Upload audio

Drop a track. The multimodal LLM listens to it and produces a Suno-ready prompt. If you already have the lyrics, paste them below for canonical wording.

⤓ Drop an audio file or click to browse — mp3, wav, flac, m4a, ogg, aac · max 500 MB

or

Paste a YouTube / audio link We'll download the audio and process it the same way.

Reference lyrics (optional but recommended) Paste lyrics here for canonical wording. The model will still add audio extras (repeats, ad-libs, backing vocals).

02

Generated prompt

All fields are editable. Char counters enforce Suno's limits.

Tempo

Key

Duration

Timbral fingerprint

librosa measurements (click to expand)

Brightness (centroid)

Rolloff

Flatness

ZCR (noisiness)

RMS loudness

Dynamic range

Onset density

Heuristic genre

Title

editable

Description

narrative paragraph(s)

Positive (styles)

0 / 500

Exclude (styles)

0 / 500

Plain comma-separated phrases for things Suno should avoid. The field itself implies negation — no -- prefix needed.

Lyrics block

0 / 5000

03

Create a variant

Reinterpret the song in a new style or language. Leave both as "Keep" to disable.

Lyrics

Style

04

Variant

Reinterpretation of the original above. All fields editable.

Title

editable

Description

editable

Positive (styles)

0 / 500

Exclude (styles)

0 / 500

Lyrics block

0 / 5000