Suno V6 Prompt: Examples, Style Tags, Settings, and Free Options
Copy-ready Suno V6 prompt examples for rock, jazz, pop and lo-fi, the style tags worth using, how the Variety slider changes results, what free accounts get with v6-mini, and free alternatives that work today.
Most bad AI songs come from one lazy sentence: "make me a cool pop song." Suno V6 rewards the opposite habit. Give it a clear genre, a mood, a vocal description and a structure, and it hands back something that sounds like a deliberate production instead of a coin flip. This article gives you the Suno V6 prompt formula, a bank of examples you can paste today, the style tags that actually shift the sound, the settings that matter, and what you can do for free.
What Suno V6 Changed
Suno's V6 release is not one model but a family of three: v6, v6-wild and v6-mini. That split matters for your prompts because each one reacts differently to the same text.
Model
Best for
Behavior
Who gets it
v6
Lyric-led songs, clear structure
Precise and predictable
Pro and Premier
v6-wild
Experiments, odd genre blends
Less predictable, more surprising
Pro and Premier
v6-mini
Fast sketches, high-volume testing
Lighter and quicker
Everyone, including free accounts
Mentions of an eight-minute maximum per generation are common across early write-ups, and one vendor test reports the lyrics box accepting around 5,000 characters and the style box around 1,000. Treat those numbers as reported, not official, and check them in the composer before you plan a long track around them.
💡 Quick rule: use v6 when your prompt includes verses, a chorus, a bridge and a vocal description. Use v6-wild when you want to be surprised. Use v6-mini to test ideas cheaply before you spend credits.
The Four-Part Prompt Formula
A strong style prompt answers four questions in one short line. Think of it as a casting call for a band.
Genre and era. "90s boom bap," "modern country pop," "early 70s soul."
Mood and energy. "Melancholic," "triumphant," "late night, slow burn."
Vocal character. "Warm female alto, close mic, breathy" or "raspy male baritone."
Instruments and production. "Fingerpicked acoustic guitar, upright bass, tape saturation."
Put them in that order, separated by commas. The most important words go first, because models tend to weight the start of a prompt more heavily.
Paste these into the style field, then add your own lyrics or let the model write them.
Indie Pop
Dreamy indie pop, bittersweet, breathy female vocal with light reverb, clean electric guitar, soft analog synth pad, tight drums, 108 BPM
Classic Rock
Late 70s hard rock, raw and sweaty, gritty male vocal, crunchy Les Paul tone, driving bass, live drums with room sound, 132 BPM
Lo-Fi Study Beat
Lo-fi hip hop, warm and sleepy, instrumental, dusty vinyl crackle, mellow Rhodes piano, muffled drums, gentle tape wobble, 78 BPM
Smoky Jazz
Mid-century jazz trio, intimate late night club, brushed snare, walking upright bass, expressive piano, warm room ambience, instrumental
Cinematic Folk
Cinematic folk, hopeful, soft male vocal with layered harmonies, fingerpicked acoustic guitar, cello, slow build to a big final chorus, 92 BPM
Modern Country
Modern country pop, sunny and nostalgic, confident female vocal with a slight twang, acoustic strum, pedal steel, handclaps, big singalong chorus
💡 Notice that none of these prompts mention an artist name. Describe the sound instead. You get cleaner results and you avoid the filters that block named artists.
Style Tags That Shift the Sound
Style tags are short descriptors you stack inside the style field. The trick is to use a handful, not forty. Here are groups that tend to produce audible differences.
One independent test reports that the style text is stored exactly as you paste it. That means punctuation and order are yours to control. If you want a tag to carry real weight, move it to the front.
Tags to Skip
Contradictions such as "calm, aggressive" in the same line. The model will average them into something bland.
Long lists of ten genres. Pick one main genre and one flavor.
Vague praise like "amazing, epic, best quality." These words add nothing the model can use.
Lyrics Structure Tags
The lyrics box is where you direct the arrangement. Square brackets act like stage directions. Reports from early testing suggest v6 responds well to explicit section labels, so label everything.
[Intro: soft piano]
[Verse 1]
Streetlights hum on an empty road
I carry every word I never told
[Pre-Chorus]
And the night keeps leaning in
[Chorus]
Take me home, take me home
Where the porch light never goes
[Bridge: stripped back, vocal only]
Say it once and let it go
[Final Chorus: big, layered harmonies]
[Outro: fade out]
A few habits that save credits:
Keep sections short. Two to four lines per verse is plenty.
Name the shift. "[Bridge: stripped back]" tells the model what changes.
Avoid walls of text. If lyrics repeat or the song ends early, shorten the structure and make each label explicit.
Write for the ear. Read the lines out loud. If you stumble, the singer will too.
Settings That Matter
Most of the interface is simple, but a few controls shape the output more than any single word.
The Variety Slider
According to Suno's help center, the Variety slider adjusts and updates your style prompts to add variety to outputs. Set it low or to zero and your tags stay intact, so you keep full control. Push it up and Suno will reinterpret your prompt, which can produce happy accidents or drift away from your plan.
Variety setting
What happens
Use it when
0 to low
Your tags are followed closely
You have a specific sound in mind
Medium
Light reinterpretation
You want polish without surprises
High
Heavier reshaping of the prompt
You are brainstorming
Pick the Right Model First
Change only one thing at a time. A fair test keeps lyrics and style fixed and swaps only the model, so you can hear what v6, v6-wild and v6-mini each do with the same input.
Vocal or Instrumental
If you want no singing, say "instrumental" in the style field and leave the lyrics empty. Otherwise the model may invent a vocal.
Free Options for Suno V6
Here is where things stand for people who do not want to pay.
On Suno itself: free accounts get v6-mini, which Suno describes as the fastest and free version of its premium v6 models. The full v6 and v6-wild are for Pro and Premier subscribers. Reports of the free plan mention around 50 credits per day, a shared queue, and no commercial rights for free outputs. Pricing for paid tiers varies between sources, so check the official pricing page before you commit.
Plan
Models
Commercial use
Best for
Free
v6-mini
No
Testing prompts, hobby listening
Pro
v6, v6-wild, v6-mini
Yes
Regular creators
Premier
v6, v6-wild, v6-mini
Yes
Heavy use
💡 Smart free workflow: write your prompt on v6-mini, keep what works, and only move to a paid model when the idea is locked in. Mini is for sketching, not final masters.
Suno V6 Is Not on PicassoIA Yet
To be direct: Suno V6 does not have a page on PicassoIA, so you cannot run it there. But the prompt habits in this article transfer to other song generators, and several are available now. For similar results today, try:
Music 2.6: full songs with vocals and custom structure tags. The closest match to how Suno works.
Lyria 3 Pro: full-length songs from Google, good for polished, cinematic output.
Music 2.5: vocal songs with a slightly lighter feel.
Paste a style line from the examples above into the Prompt field. It accepts up to 2,000 characters, so you have plenty of room. Use the same four-part formula: genre, mood, vocal, instruments.
Step 3: Add Lyrics With Structure Tags
In the Lyrics field, type your words with tags such as [Intro], [Verse], [Pre Chorus], [Chorus], [Bridge], [Solo] and [Outro]. The field takes up to 3,500 characters. Separate lines with line breaks.
Step 4: Choose the Options
Is Instrumental: turn it on for a track without vocals. The prompt is required and lyrics are ignored.
Lyrics Optimizer: turn it on if you leave the lyrics empty and want the model to write them from your prompt.
Audio format: MP3 for quick listening, WAV when you want lossless.
Sample rate and bitrate: leave them at 44100 Hz and 256 kbps for the highest quality.
Step 5: Generate and Iterate
Run it, listen, then change one thing. Swap the vocal description, adjust the tempo, or shorten a section. Small edits beat full rewrites.
Five Mistakes That Waste Credits
Stuffing the prompt. Ten genres in one line produce mush. Two is the sweet spot.
Skipping the vocal description. Without it, you get whatever voice the model feels like.
No section labels. Unlabeled lyrics are where early endings and repeated lines come from.
Changing everything at once. If you edit genre, tempo and lyrics together, you will never know what helped.
Naming real artists. It can trigger filters and rarely gives the result you imagine. Describe the sound.
Prompt Recipes by Goal
Goal
Prompt direction
Podcast intro
Instrumental, upbeat, short, clean synth and claps, 120 BPM
Instrumental, cinematic ambient, slow build, no drums for first 20 seconds
Workout track
High energy electronic pop, punchy kick, anthemic chorus, 140 BPM
Wedding first dance
Tender acoustic ballad, intimate male or female vocal, strings, 70 BPM
A Simple Testing Routine
Treat every song idea as a short experiment:
Write one four-part style line.
Run it on v6-mini, or on Music 2.6 if you want a free path off Suno.
Listen once without judging, then once with notes.
Change a single variable: vocal, tempo or one instrument.
Save the version you like and note the prompt that made it.
After ten rounds you will have a personal bank of prompts that behave the way you expect. That bank is worth more than any list of generic tags.
Try Your Own Song on PicassoIA
You do not need a studio or a band to hear your idea played back. Pick a style line from this article, write four short lyric lines with structure tags, and run it through Music 2.6 or Lyria 3 Pro on PicassoIA. Try an instrumental lo-fi loop first, then a full vocal track with a bridge. Compare the results, tweak one detail at a time, and keep the versions that surprise you. Your next favorite song might be one prompt away.