Tips for Better AI Prompts

The quality of your AI song is strongly influenced by how you write your prompt.
Clear and well-structured prompts help OpenMusic understand your intent, resulting in music that feels more accurate and intentional.

Include the key elements

A strong prompt usually covers these six areas:

  • Genre / Style
  • Instruments
  • Vocals
  • Tempo
  • Mood
  • Production / Mix

You don’t need to overthink it—just make sure each element is mentioned clearly.

Example prompts

Basic: “Pop song”

Improved: “Upbeat pop song with female vocals and acoustic guitar”

More detailed: “Upbeat indie pop with bright acoustic guitar, warm female vocals, around 120 BPM, cheerful mood, clean modern production”

Each added detail gives the AI more guidance and reduces randomness.

Be specific about genre and style

Avoid using only broad labels like “pop”, “rock”, or “electronic”.
Specific subgenres and stylistic hints lead to more consistent results.

Good genre phrases

  • “Synth-pop with 80s vibes”
  • “Lo-fi hip-hop for studying”
  • “Cinematic orchestral trailer music”
  • “Melodic techno for late-night club atmosphere”

Example prompt

“Lo-fi hip-hop beat with soft piano and vinyl crackle, relaxed late-night study vibe”

This tells the AI not only the genre, but also the purpose and feeling.

Describe instruments clearly

Instead of listing instruments, describe how they sound and interact.

Good instrument phrases

  • “Jangly electric guitar with light overdrive”
  • “Warm acoustic guitar fingerpicking”
  • “Punchy drums with strong snare and deep kick”
  • “Gentle piano arpeggios with soft synth pads”

Example prompt

“Indie rock track with jangly electric guitars, punchy drums, melodic bass, and subtle synth pads in the background”

Clear descriptions help balance the arrangement correctly.

Give clear vocal direction

If your song includes vocals, describe them clearly.
This includes gender, tone, emotion, and delivery style.

Useful vocal descriptors

  • Gender: “female”, “male”, “androgynous”
  • Tone: “warm”, “bright”, “raspy”, “dark”
  • Emotion: “melancholic”, “joyful”, “tender”, “confident”
  • Delivery: “intimate”, “powerful”, “laid-back”, “whispered”

Example prompts

“Warm female vocal, intimate and slightly breathy, emotional chorus, calm verses”

“Raspy male vocal, energetic and powerful, suitable for a rock anthem”

Set tempo, rhythm, and mood

You can specify tempo numerically or describe how the rhythm should feel.

Tempo and rhythm examples

  • “Around 120 BPM, steady four-on-the-floor dance beat”
  • “Slow ballad around 70 BPM, gentle groove”
  • “Mid-tempo track with a relaxed, laid-back rhythm”

Mood and atmosphere examples

  • “Nostalgic but hopeful, like a summer evening”
  • “Dark and tense, similar to a thriller intro”
  • “Dreamy and peaceful, suitable for meditation”

Example prompt

“Mid-tempo indie pop, around 118 BPM, nostalgic but hopeful mood, music for a sunset drive”

Mood descriptions often have a strong impact on the final result.

Add simple production notes

Production hints tell OpenMusic how polished or raw the song should feel.

Useful production phrases

  • “Clean, radio-ready pop mix”
  • “Raw live band sound with minimal processing”
  • “Lo-fi texture with gentle noise and warmth”
  • “Spacious cinematic mix with wide reverb”

Example prompt

“Emotional piano ballad with female vocal, slow tempo, intimate mood, clean and polished production with the vocal upfront”

Avoid common mistakes

Some prompts reduce quality by being vague or conflicting.

Too vague: “Cool song, a bit emotional”

Better: “Emotional pop ballad with piano and strings, slow tempo, melancholic but hopeful mood”

Too conflicting: “Very calm meditation music with aggressive screaming vocals”

Better: “Calm ambient music with soft pads and gentle vocals, relaxing and peaceful”

Clear intent always beats clever wording.

Improve results step by step

A simple iteration workflow works best:

  1. Write a clear prompt and generate a song.
  2. Listen carefully and note what feels off.
  3. Change one or two elements only.
  4. Generate again and compare.

Example refinement process

First attempt: “Happy pop song with guitar”

After listening: “Upbeat pop song with bright acoustic guitar and catchy synth lead, female vocal, around 120 BPM, joyful and energetic mood, clean radio-ready mix”

Small, targeted changes usually lead to better results than rewriting everything from scratch.