How to Use Suno’s Advanced Sliders (Weirdness, Style & Audio Influence)
Gary WhittakerFind Your Sound · LEARN · Suno implementation
How to Use Suno’s Advanced Sliders
Weirdness, Style Influence, Audio Influence and v6 Variety answer four different production questions: how far the model may explore, how strongly written style direction should lead, how strongly source audio should anchor a supported workflow, and whether Suno may vary the style prompt itself.
Page updated September 13, 2026. Platform features and availability can change.
Choose your next action
Do not read every slider page in sequence unless you need to.
Use the resource that matches the problem you are trying to solve.
Need usable starting ranges?
Choose a goal such as stable hooks, stronger genre adherence or uploaded melody preservation.
6 Goal-Based Slider RecipesNeed to prove what a setting changed?
Run a one-variable experiment and save the evidence in a Controlled Parameter Test Log.
Controlled Parameter TestingNeed to interpret the result?
Separate slider tests from v6/v6-wild/v6-mini model tests and judge what the comparison can actually prove.
Interpret a Valid Slider TestNot sure a slider is the right fix?
Compare prompting, parameter changes, regeneration and editing before spending more generations.
Intervention Selection ToolGeneral principle first
Four control jobs matter in the current v6 workflow.
Before memorizing a platform control, identify the production decision underneath it. That makes the workflow more portable when a tool changes its interface or you move to another generation system.
Suno implementation
What do Suno’s advanced sliders control?
These are influence and variation controls, not quality controls. Moving a slider higher does not automatically improve the result.
Weirdness
Use it to test how conventional or unexpected the generation may become. Lower it when the hook or arrangement needs more stability. Raise it when results feel generic and you want more unusual choices.
Style Influence
Use it to test how strongly the Styles field shapes the output. Raise it when Suno repeatedly leaves a coherent style lane. Lower it when the description is being followed too rigidly or results feel overly alike.
Audio Influence
Use it when a supported audio-reference workflow gives you an Audio Influence control. Raise it when the source melody, rhythm, phrasing or character should matter more; lower it when you want more reinterpretation.
Variety
Use it when you want Suno to introduce variation by adjusting and updating the style prompt. Suno says Variety at 0 retains full control of your style tags. For a controlled test where the written style must remain fixed, 0 is the cleanest baseline.
The part most guides miss
The controls interact with prompts, sources, models and each other.
A slider cannot be judged in isolation. Lyrics, Styles text, uploaded audio, selected model and generative variance all contribute to what you hear.
| Combination | Useful expectation | Watch for |
|---|---|---|
| High Style + low Weirdness | More conservative exploration inside a strongly signaled style lane. | Predictable or generic results. |
| High Style + high Weirdness | Unusual choices still receiving a strong style signal. | Confused arrangements when the brief is overloaded. |
| High Audio + high Style | Source audio and written direction both exert strong pressure. | Conflicting inputs pulling in different directions. |
| Variety above 0 + Style test | More variation may enter through the style prompt itself. | A muddier experiment because the written style input is no longer held as tightly fixed. |
| v6-wild + higher Weirdness or Variety | Multiple sources of deliberate unpredictability may combine. | Extra drift that cannot be cleanly attributed to one control. |
Starting points, not presets
Where should you start?
There is no universal best setting. The numeric ranges below are Jack Righteous workflow guidance, not official Suno presets. Use them as broad starting points, then change one variable.
| Goal | Weirdness | Style | Audio | Variety | Why |
|---|---|---|---|---|---|
| Controlled baseline | 45–55 | 50–65 | 50–65 if used | 0 | Keeps style tags fixed while establishing a readable baseline. |
| Stable hook or chorus | 25–40 | 65–80 | Only if relevant | Start at 0 | Reduces experimentation while reinforcing a clear lane. |
| Clearer genre adherence | 35–50 | 70–85 | Optional | Start at 0 | Lets you judge the written style before adding prompt variation. |
| Preserve uploaded idea | 30–50 | 50–70 | 65–85 | Start at 0 | Lets source and written direction establish the baseline first. |
| Explore a creative detour | 60–75 | 40–65 | 30–55 if used | Raise from 0 deliberately | Opens exploration while making prompt variation an intentional variable. |
Avoid fake precision. A value of 72 is not inherently wiser than 70. For Variety, do not invent a magic number: 0 is useful when you need style tags held fixed; otherwise raise it deliberately and judge the result.
Open the 6 Goal-Based RecipesApply the principle
The one-variable v6 slider sequence
Keep lyrics, Styles text, source audio, model and all non-target controls stable. Save each result before changing anything. If Audio Influence is unavailable, skip that pass rather than forcing a five-generation ritual.
Build the baseline
Record the model and all control states. Use Variety 0 when you need the style tags held fixed.
Test Weirdness
Move only Weirdness. Did the result become usefully safer or more surprising?
Test Style
Reset Weirdness, then adjust Style Influence. Did the intended lane become clearer?
Test Audio
If an upload is involved, reset the earlier controls and test Audio Influence. Did the right source traits move closer or farther?
Test Variety
Reset the earlier controls, then raise Variety from 0. Did variation help, or did changing the style-prompt input weaken control?
Then run a separate model test if needed
Hold the prompt, source and slider states fixed and compare v6, v6-wild and v6-mini as the variable. Do not treat the paid-access model name as the winner in advance.
Listen for five things
- Hook: memorable and repeatable?
- Style: recognizably in the requested lane?
- Structure: sections develop coherently?
- Source: did the right details survive?
- Artifacts: distortion, awkward phrasing or clutter?
Troubleshooting
Fix the symptom, not every setting.
| Symptom | First move |
|---|---|
| Suno ignores my style | Simplify the Styles field to one coherent lane, set Variety to 0 for the test, then try a moderate Style Influence increase. |
| Every result sounds generic | Add one distinctive musical detail and test a small Weirdness increase before changing multiple controls. |
| The song becomes chaotic | Lower Weirdness. Also confirm you did not switch to v6-wild or raise Variety at the same time. |
| My style tags seem to shift | Set Variety to 0 when you need full control of the written style tags, then retest. |
| The upload is barely present | Use the cleanest relevant excerpt and test Audio Influence upward gradually. |
| The upload dominates | Lower Audio Influence and make the intended transformation explicit in Styles. |
| Nothing I change is measurable | Restore one baseline, lock the model, and move only one control by a meaningful amount. |
Where to go next
Follow the problem you actually have.
Build evidence
Complete one real slider test and leave with a saved or exported record.
Controlled Parameter TestingNeed the bigger creator map?
Use the Creator Starter Kit when you need to connect this technical decision to the project, rights, positioning and the next practical move.
Creator Starter KitBuild a repeatable workflow
Connect prompting, structure, references, parameter tests, revision and song decisions into one reusable operating system.
Find Your Sound — Operator ManualPrefer to keep learning free?
Continue through the public core learning road before choosing a paid system.
Find Your Sound — Free Core PathNeed deeper training or support?
Review current training access after you know what depth of implementation you need.
See current accessQuick answers
Suno slider FAQ
What is the best Weirdness setting in Suno?
There is no universal best value. Start from a range appropriate to the musical goal, then compare generations with the other meaningful inputs held as stable as practical.
What does Style Influence mean in Suno?
It changes how strongly the information in the Styles field influences the generation.
What does Audio Influence do in Suno?
When the control is available in a supported audio-reference workflow, it changes how strongly source audio anchors the new generation.
What does Variety do in Suno v6?
Suno says Variety introduces variation by adjusting and updating style prompts. Setting Variety to 0 retains full control of your style tags.
Is Variety the same as Weirdness?
No. Weirdness is an exploration control. Variety can change the style-prompt input itself, so treat it as a separate variable.
Should Variety always be 0?
No. Zero is useful for controlled comparisons or when you want the style tags held fixed. Raise it deliberately when prompt-level variation is part of the creative goal.
Do v6, v6-wild and v6-mini make the same slider values equivalent?
No. The selected model is another variable. Hold sliders and inputs fixed if you want to compare models fairly.
Can sliders make Suno copy audio exactly?
No. Influence controls shape a new generation; they do not guarantee exact reconstruction.
Official platform references
Sources used for Suno-specific definitions
Feature names, availability and behavior can change. The testing ranges on this page are Jack Righteous workflow guidance, not official Suno presets or guarantees.
Suno Help · How to Use Creative SlidersSuno Help · v6 FAQ and VarietySuno Help · What’s New in v6Suno Help · Current v6 ModelsSuno Help · Remaster and Variation StrengthCreator control, not button memorization
Keep the method even when the interface changes.
A useful AI music workflow is not “move sliders until something works.” Decide what may vary, what must stay constrained and what must be preserved—then run the smallest test that can answer the question.
Run the free test labBuild the full systemBrowse Suno guides