The Beginner's Guide to Suno v6 prompting: Production Notes and Style Tags
By the end, you will have a generated song draft, a reusable prompt, and a comparison log. This Suno v6 prompting tutorial separates production notes, style tags, and lyrics into practical inputs. It assumes you can copy text, navigate a browser, and listen through headphones. No music theory, recording equipment, or audio plugins are required.
1. What you need
- A Suno Pro or Premier account with available generation credits for the standard v6 model.
- Suno’s browser interface at
suno.com/create, with Custom or Advanced input available. - Headphones or speakers, with playback volume kept unchanged during comparisons.
- A plain-text editor and a new file named
late-train-prompt-log.txt. - The original practice lyrics supplied below, or your own lyrics with matching section lengths.
2. Step 1 — Set up Suno v6 prompting controls
Open suno.com/create and sign in. Open the model picker at the Create form’s upper-right corner. Select v6, not v6-wild or v6-mini.
Suno’s current model documentation lists v6 under Pro and Premier access. Free accounts have v6-mini access instead. If only v6-mini appears, select it and record that substitution in your log. Do not combine results from different models in the same comparison.
Switch from Simple to Custom, or the detailed Advanced form shown in your interface. Locate Lyrics, Style or Styles, and Title. Turn Instrumental off because this exercise includes a lead vocal.
Leave reference audio, Voices, and other optional inputs empty. Those inputs would introduce additional influences during testing.
Enter Late Train A in Title. Add Model: v6 to your text file, changing the value if necessary.
The setup passes when separate lyric and style fields appear, with your chosen model displayed.
Steps, as measured in this article's example.
3. Step 2 — Write the production brief outside Suno
In your text file, paste the following brief. Keep it outside the Lyrics field.
Project: Late Train.
Genre: Synth-pop.
Tempo target: 100 BPM.
Meter target: 4/4.
Mood: Reflective, with a hopeful chorus.
Voice: Solo low-register female lead, clear diction, restrained delivery.
Rhythm: Steady electronic kick and snare.
Bass: Rounded synth bass.
Harmony: Warm sustained synth chords.
Verse production: Sparse, intimate, dry lead vocal.
Chorus production: Wider synth chords, doubled lead vocal.
Ending: Final chord resolves without a fade.
Treat these as production notes, not additional song lyrics. They describe what each musical element should do.
For this Suno v6 prompting exercise, keep one genre and one tempo target. Do not add acoustic folk, orchestral scoring, or another competing production direction.
The tempo and meter are requests, not guaranteed output properties. Likewise, a vocal description does not establish a persistent singer identity.
The check passes when the brief assigns clear roles without contradictory instructions.
6. Step 5 — Set exclusions and generation controls
Open Advanced Options, or Controls if that label appears. Find Exclude or Exclude Styles, when available, and enter this list.
rap, screamed vocals, distorted guitars
Do not place the excluded terms in the main style text. Keep the positive description and the exclusion list separate.
Set Variety to 0, also displayed as Off in some interfaces. Suno’s v6 FAQ says Variety can adjust and update style prompts. Its zero setting preserves greater control over supplied style tags.
If the sliders appear, set Weirdness to 20% and Style Influence to 75%. These are working values for this exercise, not universal settings. Suno describes Weirdness as a Safe-to-Chaos range and Style Influence as Loose-to-Strong adherence.
Leave Max Mode off if displayed. Leave other controls unchanged and record their visible values.
This Suno v6 prompting setup passes when the exclusions and available slider values match your log.
7. Step 6 — Generate and score the baseline
Click Create once. Wait until the returned tracks finish generating before changing any inputs.
Label the returned candidates A1, A2, and so forth in your log. Copy each candidate’s song URL from the browser address bar after opening it. These links identify the audio without depending on an editable title.
Play each candidate from its beginning. Listen at the same comfortable playback level throughout the comparison.
Score each item below with 0, 1, or 2. Give zero for absent, one for partial, and two for clearly present.
- Voice: One restrained, low-register female lead dominates the song.
- Rhythm: The electronic groove remains steady through the chorus.
- Arrangement: The chorus sounds wider or fuller than the verse.
- Lyrics: The supplied words remain intelligible, without sung instructions.
- Ending: The final phrase resolves without an abrupt cutoff.
Add the five values for a total out of 10. This is an exercise score, not a general quality measurement.
For Suno v6 prompting, log evidence alongside numbers. Write chorus remains narrow rather than production feels wrong.
The baseline passes when every returned candidate has a URL, score, and one concrete listening note.
8. Step 7 — Test one production-note change
Choose the baseline candidate with the clearest lyrics. Resolve ties by choosing the higher total score.
Return to the Create form and confirm your original inputs remain intact. Change Title to Late Train B.
In the style field, replace only this sentence.
Choruses widen through broader synth chords and doubled lead vocals.
Enter this replacement.
Choruses widen through broader synth chords, doubled lead vocals, and a bright upper-register synth melody.
Leave the lyrics, exclusions, model, and sliders unchanged. Save the revised prompt as Style B, then click Create once.
Score the new candidates with the same five-item checklist. Add one extra observation: Upper synth melody: absent, partial, or clear.
A fresh generation can change melodies and performances beyond your edited instruction. This Suno v6 prompting comparison tests a tendency, not an isolated audio effect.
The change is audible when a distinct high synth line appears around the chorus vocal.
9. Step 8 — Test one local arrangement cue
Keep Style B unchanged for this test. Change Title to Late Train C.
In Lyrics, change the bridge heading from [Bridge] to the following heading.
[Bridge: sparse synth chords, drums drop out]
Leave both bridge lyric lines untouched. Do not add a separate paragraph saying production notes or invent channel-control syntax.
This bracketed wording is an arrangement request, not a documented mixer command. The generation may interpret, ignore, or occasionally sing the instruction.
Save the revised lyrics as Lyrics C in your log. Click Create once, then locate the bridge using its opening words: One more light beyond the rain.
Record whether drums continue, reduce, or disappear during those two lines. Also record whether the chorus restores the fuller rhythm afterward.
For Suno v6 prompting, keep the cue only if it produces a useful musical change without unwanted sung text.
The test passes when your log identifies the bridge behavior and any instruction-reading artifacts.
10. Step 9 — Select a draft and save its prompt package
Compare your highest-scoring A, B, and C candidates. Replay the verse-to-chorus transition and bridge in each candidate.
Keep playback volume unchanged. Do not award extra points because one track starts louder.
Choose one candidate using the original ten-point checklist. If scores tie, select the version with clearer words. Record Selected candidate, its song URL, and its score.
Now assemble this package in late-train-prompt-log.txt.
Selected candidate: Paste the selected song URL.
Model: Enter the actual model.
Title: Enter the selected generation title.
Style: Paste the exact style text for that candidate.
Lyrics: Paste the exact lyrics for that candidate.
Exclude: rap, screamed vocals, distorted guitars.
Variety: 0, if available.
Weirdness: 20%, if available.
Style Influence: 75%, if available.
Unchanged controls: Record their displayed values.
Listening score: Enter the total out of 10.
Known issue: Enter one remaining problem, or none heard.
Replace every placeholder with your actual information. Keep the selected song available in your account’s library.
This Suno v6 prompting package preserves your inputs, not a guaranteed repeat of the same performance.
The workflow is complete when the saved URL opens your selected draft and its matching prompt package is complete.
11. Troubleshooting
- Step 1 shows only v6-mini. Confirm your account plan and model picker. Continue with v6-mini if needed, but record it throughout the test. Restart comparisons if you later switch models.
- Step 5 changes the style wording. Return Variety to 0 or Off. Restore
Style Afrom your text file before generating again. Preserve the rewritten version separately if it contains useful wording. - Step 6 returns an overly busy verse. Replace
Sparse intimate verses with a dry lead vocal.withVerses contain only lead vocal, synth chords, and minimal drums.Keep every other input unchanged for the next test. - Step 7 adds no audible upper synth melody. Confirm that the replacement sentence reached the style field. Generate another candidate with identical settings before revising additional instructions. Log continued absence rather than assuming guaranteed compliance.
- Step 8 sings the bridge instruction. Restore the heading to
[Bridge]. AddThe bridge drops drums and retains sparse synth chords.to the style field instead. Generate again and check both bridge lines for spoken or sung directions.
12. Tools
Valkat Studio covers lyrics, Suno prompts, vocal profiles, mastering, cover art, and music videos. Its audio-finishing tools are grouped in the mastering suite.
13. Final thoughts
Keep the selected draft, its exact inputs, and one unresolved issue together. Future Suno v6 prompting tests should change one instruction before creating another comparison. Treat production notes and style tags as musical guidance, then verify their effect by listening.