For my first 50 songs in Suno, I typed things like "acoustic song about love" and hit generate. The results were fine. Generic. Forgettable. The kind of music you'd hear in a waiting room and not notice.

Then I figured out the formula, and everything changed.

The Six-Part Formula

Every prompt I write follows the same structure: Genre + Vocals + Lead Instrument + Production Style + Mood + Tempo.

That's it. Six specific descriptors. Four to seven words total in each category. Here's what each one does and why it matters.

Genre sets the foundation. But "pop" is too broad. "Indie folk singer-songwriter" is specific. "Contemporary R&B" is specific. "Pop-punk with emo influence" is specific. The more precise your genre call, the more Suno knows what sonic world to operate in.

Vocals shape the most human element of the song. "Male vocals" gives you something random. "Warm male vocals, clear and intimate" gives you a specific person. Other tested vocal descriptors that produce distinct results: "raspy," "breathy," "powerful," "gritty," "smooth," "falsetto," "belting," "conversational." Each word changes the voice Suno generates.

Lead Instrument tells Suno what drives the song. "Acoustic guitar driven" produces a fundamentally different track than "piano-led" or "synth-forward." Be specific about the instrument - "fingerpicked nylon string" sounds nothing like "strummed steel string acoustic." Both are guitar. Both sound completely different.

Production Style controls the recording feel. "Polished pop production" sounds like a radio single. "Raw and intimate recording" sounds like a voice memo from someone's bedroom. "Lo-fi production, vinyl crackle" sounds like it was recorded in 1997. These descriptors tell Suno how much to fill the sonic space and how clean to make everything.

Mood is the emotional instruction. "Uplifting and hopeful" versus "melancholic and introspective" - same genre, same instruments, completely different energy. Mood is where you tell Suno how the song should make someone feel.

Tempo locks the speed. Without it, Suno picks whatever it wants - which might be perfect or might be 40 BPM off from what you imagined. Adding "110-120 BPM" or even just "moderate tempo" gives you control.

The Formula in Action

Here are five prompts I actually use, with what each one produces:

Prompt 1: "Indie folk singer-songwriter, warm male vocals, acoustic guitar driven, light percussion, uplifting and hopeful, modern production, clear and intimate vocal recording, gentle piano accents, 110-120 BPM"

What it produces: Clean, warm, guitar-forward songs. The kind of thing that shows up on a "Coffee Shop Acoustic" playlist. Consistent vocal character with emotional range.

Prompt 2: "Pop-rock, powerful female vocals, electric guitar driven, punchy drums, modern production, radio-ready, energetic and emotional, bass-heavy chorus, 120-140 BPM"

What it produces: High-energy, guitar-forward rock with pop hooks. Confident female vocal that shifts between verse vulnerability and chorus power.

Prompt 3: "Acoustic ballad, warm emotional male vocals, fingerpicked guitar, soft piano, intimate and reflective, minimal production, spacious mix, strings in chorus, 70-85 BPM"

What it produces: Slow, stripped-back emotional songs. Vocal sits right in front of you. Piano and strings swell in the chorus. The kind of song that makes you stare out a window.

Prompt 4: "Children's music, playful and energetic, bright acoustic instruments, cheerful female vocals, educational and fun, singalong melody, ukulele and percussion, 110-130 BPM"

What it produces: Catchy, bouncy kids songs that adults don't hate listening to. Clear vocals, simple arrangements, hooks that stick.

Prompt 5: "Contemporary R&B, smooth male vocals, lush production, synth pads, subtle beat, sensual and atmospheric, falsetto, 80-95 BPM"

What it produces: Moody, smooth R&B with layered production. Vocal floats over atmospheric synths. Nighttime energy.

What Happens When You Drop a Category

I tested this. Same lyrics, same basic intent, but removing one category from the prompt each time.

Without genre specificity: the song defaulted to generic pop. Not bad, but not distinctive.

Without vocal descriptors: Suno picked a random voice that didn't match the genre feel. Got a bright, thin voice on what should have been a warm ballad.

Without instrument calls: production felt aimless. No clear sonic identity. Everything was there and nothing stood out.

Without production style: technically fine but lacked character. Like a well-recorded song with no personality.

Without mood: energy was flat. The song didn't know if it was happy or sad, so it landed in the middle - which is nowhere.

Without tempo: got a ballad when I wanted mid-tempo. One word would have prevented a wasted generation.

Every category does real work. Drop any one and the output degrades noticeably.

Words Suno Actually Responds To

Through testing hundreds of prompts, I've found that certain words produce dramatic differences and others are basically ignored.

High-impact words (use these):

Warm, intimate, raw, polished, atmospheric, driving, gentle, powerful, gritty, spacious, lush, minimal, building, anthemic, lo-fi, crisp, dreamy, punchy, layered

Low-impact words (don't waste prompt space):

Professional, high-quality, good, amazing, beautiful, perfect, best. These are evaluations, not descriptions. Suno doesn't know what "good" means. It knows what "warm" means.

Words that change vocals specifically:

Breathy, raspy, smooth, clear, rich, thin, belting, falsetto, conversational, passionate, tender, gritty, soulful

Words that change production specifically:

Spacious, dense, compressed, dynamic, vintage, modern, organic, synthetic, layered, stripped, polished, raw

The Advanced Move: Prompt Stacking

Once you have the formula down, you can stack modifiers for precision.

"Warm male vocals" is good. "Warm male vocals, clear and intimate, occasional falsetto" is better. You're giving Suno three data points about the voice instead of one.

"Acoustic guitar" is good. "Fingerpicked acoustic guitar, nylon string, close-miked" is better. Three data points about one instrument.

The limit is around 7-9 total descriptors before Suno starts getting confused. So spend your budget wisely - stack detail on the elements that matter most for your song, and keep the others simpler.

Copy These Prompts

If you want to start producing better music immediately, copy any of the five prompts above, paste your own lyrics, and generate. These are tested. They work.

Then start modifying. Swap "warm" for "raspy." Change "acoustic guitar" to "piano." Add "lo-fi" to the production. Each change teaches you what Suno responds to. After 10-15 experiments, you'll have your own intuition for what works.

And if you want the complete prompt library across every major genre - with tested variations, troubleshooting guides, and the framework for building your own - that's the core of what I teach.


I've tested hundreds of prompt combinations across eight genres. The patterns are clear. The formula works. If you want the full tested library, I'm packaging everything into a course with downloadable Claude AI skill files that automate the prompt-building process.