All articles

Suno V6 Features: Everything New, Everything Removed, and What It Means for Your Music

Suno shipped its biggest update ever on September 9, 2026, and it did two things at once: it gave us the most capable set of Suno V6 features we have ever had, and it deleted every model that came before. If you make music with Suno, both halves of that sentence matter equally.

Here is the short version before we go deep. V6 is a family of three models, not one. The flagship v6 and the experimental v6-wild are for Pro and Premier subscribers, while v6-mini is free for everyone. The headline features are plain language editing, image and video references, a Max Mode toggle for consistency on longer songs, and a new Variety slider that quietly rewrites your style prompts unless you turn it off. Underneath all of that sits the real story: V6 is the first Suno generation built with licensed music from Warner Music Group, BMG, and Believe, which is why the old models had to go.

I have spent the days since launch generating with all three models, digging through Suno’s own documentation, and reading what other producers are finding. This article covers every confirmed feature, the ones Suno buried in the FAQ, the community discoveries that are not in any official post, and the honest downsides. Because there are downsides.

Three Models Instead of One

For the first time, Suno is running three models side by side, and picking the right one for the job is now part of the craft.

According to Suno’s official release notes, v6 is the flagship. Suno describes it as reliable and precise, built to deliver polished output across genres. In my testing that description holds up. When I feed it a detailed style prompt with specific instrumentation, it follows instructions more faithfully than v5.5 did, and it wanders off less. If you know exactly what you want, this is your model.

v6-wild is the interesting one. Suno positions it as the exploration model, less predictable and more varied, producing what they call unexpected, textured, and ambitious results. Translation from marketing to musician: it takes bigger swings and misses more often. When it hits, the results have character that the polished flagship sometimes lacks. I will come back to v6-wild later, because it has become the unofficial answer to a question a lot of upset users are asking.

v6-mini is the free tier model, described as a faster, more efficient version of the premium models. It is genuinely usable for sketching ideas, but be clear eyed about its limits. As Digital Music News reported at launch, free accounts get v6-mini without downloads or commercial rights. It is a demo environment, not a production tool. If you plan to release anything, mini is for auditioning ideas before you spend paid credits, nothing more.

Both v6 and v6-wild require a Pro or Premier subscription. There is no way to touch the full models on a free account.

Why the Licensing Shift Matters More Than Any Feature

Before we get into the fun toys, you need to understand what V6 actually is under the hood, because it changes the legal picture for everyone who releases AI music.

V6 is the first Suno generation developed in partnership with major rights holders. According to Music Business Worldwide’s launch coverage, Suno built the v6 family with Warner Music Group, BMG, and Believe, and the company says the new models were not trained on the data used for earlier versions. Suno’s Chief Product Officer Jack Brody confirmed that Universal Music Group and Sony Music recordings were not part of the new training data, and that Suno is sharing a portion of its revenue with its licensing partners.

This is why every previous model got retired on launch day. The old models are the ones at the center of the copyright lawsuits from Universal, Sony, and others, and those cases are still active. A Munich court ruled against Suno in a case brought by German collecting society GEMA at the end of July, and Suno says it may appeal. The V6 launch does not settle any of that. What it does is move the product you are actually using onto cleaner legal ground.

For those of us who distribute music, this matters in a practical way. Distributors have spent two years treating Suno output as radioactive. Tracks made on a licensed model, with label partners collecting a revenue share, are a different conversation. Believe owns TuneCore, and it joined the partnership the day before launch, which is not a coincidence. Early community reporting suggests TuneCore is accepting partner model tracks again, though I would still expect scrutiny from other distributors. Do not assume the licensing deal is a universal green light. It is not.

Natural Language Editing Is the Feature That Changes Daily Work

Of all the new Suno V6 features, this is the one I use most. You can now edit part of an existing song by describing the change in plain text, and the model changes that part while leaving the rest alone.

Suno’s own examples from the v6 FAQ show the range. You can ask it to change the chorus so it is sung by a gospel choir. You can update a single lyric, swapping one word for another, without regenerating anything else. You can ask for the intro to be shorter. Before V6, most of these jobs meant rolling the dice on a full regeneration and hoping the parts you loved survived. They usually did not.

In practice, the section level edits work better than I expected and the surgical lyric swaps work better still. Changing one word genuinely leaves the rest of the vocal intact most of the time. Bigger structural requests are less reliable. When I asked for a completely different bridge with new chords, the model sometimes altered the surrounding sections too, so listen to the whole track after every edit, not just the part you changed.

One workflow note: the classic tools did not disappear. Cover, Extend, Crop, and Replace Section still exist under the Remix menu on each track, now powered by V6. The plain language box is a new front door to those tools, not a replacement for them.

Image to Song and Video to Song

V6 accepts images and video clips as creative references alongside text and audio. Drag in a photo, and the model reads the mood of it and uses that as a starting point. Feed it a video clip, and it generates music shaped by what is happening on screen.

Suno’s framing is deliberately loose. Their example prompt is “make a song based on this image, this audio, and my journal entry,” and the FAQ confirms you can stack multiple reference types, including Suno songs, playlists, audio uploads, images, and video, in a single request. The vibe based prompting works the same way: “make a song that feels like midnight on a rooftop” is an official example, and V6 handles that kind of impressionistic brief far better than any previous version.

My honest read after testing: image references are a mood board, not a spec sheet. A photo of a rainy street reliably nudges the output toward something moody and downtempo, but two generations from the same image can land in very different genres unless your text prompt pins the style down. The best results come from giving each input a job in the text, something like “the image sets the atmosphere, the audio clip is the tempo reference.”

Video to song is the more practically useful of the two if you score content. It effectively turns any clip into a scoring brief, which is a real workflow for YouTube and short form creators. Two warnings, though. First, Suno confirms that feeding many images and videos into a prompt increases the credit cost of the generation, so multimodal experiments are not free. Second, only upload visual material you have the rights to use as a reference. Dropping in a clip from a film you do not own is asking for trouble you do not need.

Max Mode: When to Pay Double

Max Mode is a simple toggle with a simple trade. Turn it on, and V6 spends more compute on your generation to keep the song consistent from start to finish. It costs more credits, roughly double the standard rate based on what the interface shows, so a generation that normally runs 10 credits for two songs runs about 20.

Suno’s FAQ is unusually specific about when it helps: songs longer than two minutes, covers where the result needs to stay close to the original, transferring the style of one song onto another, and keeping vocals and style consistent through the whole track. For quick ideas and shorter songs, Suno says standard mode is all you need, and I agree.

The two minute threshold is the detail to internalize. Multiple users have reported audio quality and vocal identity drifting after roughly the two and a half minute mark on standard mode, and my own longer generations back that up. The singer at 3:10 is sometimes not quite the singer from the first verse. Max Mode largely fixes this.

My rule after a week: explore on standard mode, commit on Max Mode. Generate your cheap pairs until one take is close to what you want, then rerun that exact configuration with Max Mode on for the keeper. Leaving Max Mode on while you are still exploring is the fastest way to burn a month of credits on drafts.

The Variety Slider Is Rewriting Your Prompts

The Variety slider is the most misunderstood control in V6, and if you write careful style prompts, you need to know what it actually does.

Suno’s FAQ says it plainly: the Variety slider introduces variety in your outputs by adjusting and updating your style prompts, and if you want to retain full control of your style tags, you should reduce it to 0. Read that again. At any setting above zero, Suno is rewriting the style prompt you typed, differently on each take. Undetectr’s breakdown of the V6 controls confirmed this behavior on camera: an engineered style box is only used verbatim when Variety is off.

The slider defaults to Normal on v6 and v6-mini, which means the default experience is Suno editing your prompts without telling you. For casual users that is probably fine, and honestly it produces pleasant surprises. For anyone who has spent months learning which style tags produce which results, it is a silent saboteur. If your carefully tuned prompt suddenly produces something unrelated, check the slider before you blame the model.

Variety also sits alongside two returning sliders that people keep confusing it with. Style Influence controls how strictly the model obeys the prompt. Weirdness controls how adventurous the output is. Variety is different from both: it changes what the prompt says before the model even reads it. Change one control at a time between generations, or you will never know which knob did what.

Mashups, Sampling, and Simple Mode

Three more V6 capabilities deserve a quick tour, because they replace whole external workflows.

Mashups let you combine elements from multiple sources in one request. Suno’s example: take the vocals from one song, the drums from another, add new lyrics, and make it 80s synthwave. This works across your own songs, which is a rights boundary, not a technical one, and I am glad Suno drew it.

Sampling and isolation now happen in a single flow. You can point at a moment in a track, something like “sample the riff at 0:45, isolate the guitar, build a beat around it,” and V6 does the extraction and the rebuild in one pass. This used to take a stem separation tool plus a DAW, a digital audio workstation, the software producers use to record and arrange music. It is now one prompt.

Simple Mode ties it together. Per Suno’s FAQ, you no longer need to know whether your request is technically a Cover, a Remix, or an Extend. Describe what you want and the model figures out the workflow. For newcomers this is a genuine improvement. For control freaks like me, the classic tools are still there when I want to be explicit.

Every Old Model Is Gone, and Your Library Is Not

Now the part that filled Reddit with cancellation threads. As V6 rolled out, Suno retired every previous model. v4.5, v4.5+, v5, and v5.5 were switched off on launch day, and there is no legacy toggle, no grace period, no opt out.

Your existing songs are safe. Suno’s FAQ confirms that everything in your library stays there unchanged, and you can still listen to, share, remaster, and cover songs made with old models, all the way back to your ancient V3 experiments. Custom models trained on v5.5 were automatically upgraded so V6 powers them going forward, and songs those custom models already made are untouched.

What you cannot do is generate anything new with an old model. Any new iteration of an old song, any extension, any cover, will come out of V6. If a chorus you love exists only as a v5.5 take, that take is now a finished recording, not a work in progress.

Whether this bothers you depends on how you used Suno. The reaction roundup at Undetectr captured both camps honestly: launch day brought a heavily upvoted “I’m cancelling” thread alongside people calling V6 the best model yet. The people hurting most are those whose specific workflow, persona, or genre only worked on one particular old model. That muscle memory is gone, and pretending otherwise would be dishonest. The licensing logic is easy to follow, since retiring the models trained on disputed data is the entire point, but the execution gave working musicians zero transition time, and that was a rough call.

Getting the Old Sound Back with v6-wild

Here is the workaround that has emerged for people who miss the old models, and it comes from two directions at once.

Officially, Suno acknowledges the problem. The FAQ says the v6 family was designed to support all the ways people liked creating with past models, and links a tutorial on how to achieve the results of your favorite past models with V6. That is corporate language for “we know V6 sounds different, here is how to steer it.”

Unofficially, the community found the answer faster: v6-wild is where the old personality went. The flagship v6 is polished in a way that can feel sanded down, especially if you loved the happy accidents of v4.5 or the rawness of earlier models. v6-wild takes the same generation and lets it get weird again. Several producers report that wild’s output feels closest in spirit to the older models, and my own experience agrees, particularly for lo fi, psych, and anything where imperfection is the aesthetic.

The practical recipe that is working for me: start on v6-wild, keep your old style tags, and set Variety to 0 so Suno stops rewriting them. If wild drifts too far into chaos, pull the Weirdness slider down below its default. When a take lands, you can bring it into the flagship v6 for refinement using the editing tools. It is not a time machine, and nothing will exactly reproduce a retired model, but it closes most of the gap.

What V6 Costs

The good news is that V6 did not come with a price hike. The plan structure is unchanged: Free at $0 with 50 credits renewing daily, Pro at $10 per month, and Premier at $30 per month, with annual billing cutting roughly 20 percent off, taking Pro to about $8 per month and Premier to about $24 per month. Prices are Suno’s US list prices; check the pricing page for your region.

Credit math is also unchanged at the base level. Suno’s FAQ confirms each generation still produces two songs for a total of 10 credits, the same as v5.5. The new costs hide in the new features. Max Mode roughly doubles a generation. Loading a prompt with multiple images and videos raises the cost too, by an amount Suno has not published precisely. Pro’s 2,500 monthly credits still go a long way if you explore on standard mode, but a Max Mode habit plus multimodal prompts will drain them noticeably faster than v5.5 ever did.

Subscription credits do not roll over month to month, so time your heavy sessions accordingly.

Downloads, Commercial Rights, and the Fine Print

Six days before V6 launched, Suno changed its terms of service, and the two changes travel together. If you release music, this section is the one to read twice.

First, downloads are now capped. As ArtistDirect summarized the new terms, effective September 3, 2026, Pro subscribers get 20 song downloads per month, Premier gets 60, and only Premier Studio users are uncapped. Free users who joined before that date have up to seven lifetime downloads, and extra downloads can be purchased. Since a download is the only way to get a track from Suno to a distributor, the cap is effectively a meter on releasing. Twenty downloads sounds like plenty until you remember stems, alternate mixes, and the takes you export just in case. Budget your downloads like you budget your credits, and only export tracks you have actually decided to keep.

Second, Suno has committed to audio watermarking and fingerprinting so songs made on the platform can be identified elsewhere, per Music Business Worldwide’s reporting on the BMG deal and the new terms. Assume every V6 file carries identification, and assume platforms will increasingly read it. If your release strategy depends on nobody knowing a track started in Suno, that strategy is dying, and I would not build on it.

Third, and most important, read what the terms actually promise about ownership. Suno assigns Pro and Premier subscribers its rights in the output you generate, but the same document states that Suno makes no representation or warranty that any copyright will vest in any output, and notes that applicable rights holders may have the right to collect revenue on distribution of outputs on third party platforms. So you can commercially use your paid tier tracks, but Suno is explicitly not promising the track is copyrightable, and label partners may take a cut downstream.

The Ownership Question V6 Does Not Solve

That last clause deserves its own section, because the licensing deal has confused a lot of people into thinking the copyright problem is fixed. It is not.

In the US, the Copyright Office’s position is that purely AI generated material is not protected by copyright, because protection requires human authorship. A license from Warner does not change that. It changes whether Suno had the right to train the model, not whether you own what comes out of it. A raw V6 export with an AI generated melody, arrangement, and vocal remains, in the US framework, largely unprotectable, no matter what plan you pay for. UK readers face a slightly different landscape, since UK law has a provision for computer generated works, but if you distribute to US platforms, the US frame is the one that bites.

Your ownership grows with your human contribution. Writing the lyrics yourself, restructuring the arrangement, adding your own performances, all of that builds a protectable layer on top of the AI foundation. And if a track genuinely matters to your catalog, having real musicians re record it from the ground up remains the cleanest path to a recording you fully own, since a human performed master is a normal copyrighted recording with none of these questions attached. For most day to day tracks that is overkill. For the one song getting a sync pitch or a serious marketing budget, it is worth thinking about.

My Take After the First Week

V6 is the best Suno has sounded and the most control we have ever had, and I say that as someone annoyed by parts of this launch. The plain language editing alone saves me more credits than any previous feature, because I fix sections instead of regenerating songs. Max Mode genuinely solves the long song consistency problem. The multimodal inputs are fun and occasionally brilliant, if not yet precise.

The things I would push back on: retiring every model overnight was needlessly abrupt, the Variety slider defaulting to on quietly undermines everyone who learned to prompt properly, and the download caps turn releasing into a metered activity right as the licensing deal made releasing more viable. Those three decisions all make sense from Suno’s side of the table. From ours, they cost trust.

If you are deciding whether to stay subscribed, my honest advice is to judge V6 on its own terms rather than as a replacement for a model you miss. Spend one session with the recipe from the v6-wild section, one session learning the edit box, and one session with Variety at 0 and your old tags. If after that the sound still is not yours, that is a real answer too.

What to Do This Week

A concrete checklist, in order. Export the old model takes you cannot live without, within your download allowance, because those recordings are now final. Set Variety to 0 in your default settings if you use engineered style prompts. Rebuild one trusted old prompt on both v6 and v6-wild and compare, changing nothing else. Save Max Mode for keepers over two minutes. And before you distribute anything, reread your distributor’s current AI policy, because the licensed model era is changing those policies month by month, mostly in our favor.

The Suno V6 features are a genuine step forward wrapped in a forced migration. Learn the new controls, protect your old work, and keep your human fingerprints all over anything you release. That combination works in every era of this tool, and it works in this one.

Sources

Common questions

What are the three Suno V6 models?

There are three. v6 is the flagship for Pro and Premier subscribers, built to be reliable and precise. v6 Wild is the experimental model on the same paid tiers, producing less predictable and more varied results. v6 Mini is the free model, faster and lighter, but without downloads or commercial rights.

Can I still use Suno v5.5 or older models?

No. Every model before V6 was retired on launch day, September 9, 2026, with no legacy option. Your existing songs stay in your library and you can still listen to, share, remaster, and cover them, but any new generation or iteration uses V6.

How does natural language editing work in Suno V6?

You describe the change in plain text and V6 edits just that part of the song while leaving the rest intact. You can change a whole section, like asking for a chorus sung by a gospel choir, or swap a single lyric word. Bigger structural edits are less reliable, so always listen back to the whole track.

What is Max Mode in Suno V6 and when should I use it?

Max Mode is a toggle that makes V6 spend more compute keeping a song consistent from start to finish, at roughly double the credit cost. Suno recommends it for songs longer than two minutes, covers, style transfers, and vocal consistency. Explore on standard mode, then use Max Mode for the take you plan to keep.

What does the Variety slider do in Suno V6?

At any setting above zero, Variety rewrites your style prompt to add variation, differently on each take. Suno confirms that setting it to 0 is the only way to keep full control of your style tags. It defaults to on, so prompt focused producers should turn it down first thing.

Can Suno V6 turn an image or video into a song?

Yes. You can drag in an image as a creative reference and V6 uses its mood as a starting point, and a video clip generates music shaped by what happens on screen. Images act more like mood boards than exact specs, and feeding many images or videos into one prompt raises the credit cost.

How can I get the old Suno sound back in V6?

v6 Wild is the closest thing. The community found that the older models' personality survives best there, and Suno published a tutorial on recreating past model results. Start on v6 Wild with your old style tags, set Variety to 0, and lower Weirdness if it drifts too far.

Does Suno V6 cost more than the old models?

No. Plans are unchanged at free, $10 per month for Pro, and $30 per month for Premier, with about 20 percent off on annual billing, and a standard generation still makes two songs for 10 credits. The new costs come from Max Mode, which roughly doubles a generation, and from heavy image and video inputs.

Was Suno V6 trained on licensed music?

Suno says the V6 family was developed with Warner Music Group, BMG, and Believe, was not trained on the data used by earlier models, and excludes Universal and Sony recordings. Lawsuits over the older models are still active, so the licensing deal improves the picture without ending the disputes.

Do I own the copyright to songs I make with Suno V6?

Not automatically. Suno assigns paid subscribers its rights in your output but explicitly does not promise any copyright will exist in it, and the US Copyright Office holds that purely AI generated material is not copyrightable. Your protection grows with your human contribution, like your own lyrics, arrangement work, or re recorded performances.