Make any song
you can imagine

Released 13 August 2026, Studio 2.0 adds seven big things: MIDI import, recording and a full piano roll; MIDI clips as prompts for audio generation; a plain-language chat bar (beta) that replaces the old Cover/Remix bar; a wavetable synth with oscillators, filters, envelopes, LFOs and a mod matrix; advanced stem separation (Split from Mix and Advanced Split); seven effects plus custom plugins you build by describing them; and drawable automation with MIDI Learn. Export is now 32-bit / 48 kHz multitrack, unlimited on Premier, and Take Lanes let you audition alternate generations without regenerating.

Instead of describing a part in words, you play it. Record or import a MIDI clip - via a hardware controller over Web MIDI, or by typing on your computer keyboard with musical typing - then use that clip as the prompt. Suno renders your performance as audio, extends it, or builds on it. Audio-to-MIDI transcription works the other way, pulling notes out of an existing part so you can reprogram it. It's the feature that turns Studio from a prompt box into a producer's tool.

A Premier subscription, and Google Chrome is recommended - Safari has no Web MIDI support, so hardware controllers won't be detected there. You'll need a desktop, laptop or tablet with at least a 768 px screen width, a CPU with SIMD support (SSE4.1 on Intel/AMD, Neon on ARM), and at least 4 GB of RAM. Uploads accept WAV, MP3 and MIDI files. Studio isn't available on phones, on the Free plan, or on Pro.

No - and it's the biggest remaining gap. Suno Studio is not compatible with VST or AU. You get seven factory effects (Compressor, Convolution, Delay, Distortion, EQ, Gate and Reverb, including sidechain compression and convolution reverb) plus any custom plugins you generate through the chat bar, which save to your library with presets, automation and MIDI Learn. If your mix depends on your own plugin folder, you'll still export stems and finish in your usual DAW.

Suno uses deep-learning models trained to turn a text description - or an audio input, or a MIDI performance - into a complete original song: lyrics, vocals, melody, instrumentation, structure, and mixing, typically in under a minute. You describe the genre, mood, and theme (or paste your own lyrics with structure tags like [Verse] and [Chorus]), and the model composes and produces the whole track. Newer versions like v5.5 add personalization: Voices for your own vocal identity, Custom models trained on your catalog, and My Taste for learned preferences.

The core loop is a single prompt box: describe your song, hit Create, and get two full takes back in seconds. Simple mode is one sentence; Custom mode adds your own lyrics, style fields, exclusions, vocal gender, and weirdness/style sliders. Around that sit a Library for your songs, a community feed for discovering and remixing others' tracks, playlists, and - on Premier - Studio 2.0, a full multitrack workstation with MIDI, effects and automation. The mobile apps mirror the web with extras like humming or recording a voice memo as input.

Songs made on the free plan are non-commercial forever, even if you upgrade later. Songs made while subscribed to Pro or Premier carry commercial-use rights: monetized YouTube, streaming releases, ads, and client work, with Suno taking 0% of revenue. But a license isn't a copyright - 100% AI-generated music may not qualify for copyright protection since it lacks human authorship, so adding your own lyrics, played MIDI parts, vocals, or arrangement strengthens your legal position. See the License section above for the full breakdown.

Generative music raises real risks - voice cloning abuse, imitating real artists, and provenance of AI content - and Suno has built guardrails for each: Voices requires identity verification (you speak a random phrase matched against your singing sample, and your voice stays private to your account), prompts referencing real artist names or real song lyrics are blocked, and generated audio carries an inaudible watermark for provenance. Remaining challenges include occasional lyric/pronunciation glitches, language mixing, and the industry litigation still working through the courts.

Content creators use it for original background music on YouTube, TikTok, and podcasts; singers demo songs in their own voice; songwriters hear lyrics performed instantly; producers play parts in over MIDI in Studio 2.0 and export 32-bit stems to their DAW; brands generate jingles and ad beds at scale; game and film composers prototype to a style; educators teach songwriting; and developers generate tracks programmatically via API platforms. And plenty of people just make songs for fun - birthday tracks, inside jokes, and gifts.

In late 2023, Suno became one of Microsoft Copilot's first creative plugins: users could ask Copilot to make a song and the Suno plugin would generate a complete track - lyrics, vocals, and instrumentals - right inside the chat. The partnership brought AI music to a mainstream audience and Microsoft sign-in remains one of Suno's login options today. It was an early signal of Suno's strategy of meeting creators inside the tools they already use.

They should be. Version 1.0 was criticised for regenerating stems using the original as a reference rather than truly separating the mix, which introduced artefacts and reverb that weren't in the source. Studio 2.0 ships dedicated Split from Mix and Advanced Split modes that isolate vocals, drums, bass, guitar, keys and effects - and they work on external audio you upload, not just Suno generations. Suno describes the fidelity as high; as this is a brand-new release, test it on your own material before committing a project to it.

Not at launch. Suno has said plugin generation requires no credits for now, but also that a credit structure may be introduced in the future - so treat it as an introductory condition rather than a permanent one. Multitrack and stem export is likewise unlimited on Premier. What still costs credits is music generation itself, including takes you listen to once and discard, so prompt carefully.

Exclude Styles is the negative prompt for music: alongside describing what you want, you list what the model should avoid - "no EDM drops," "no autotune," "exclude trap hi-hats." It's available in Custom/Advanced mode and is one of the fastest ways to fix a generation that keeps drifting into the wrong territory. Pair it with the weirdness and style-influence sliders to steer results precisely instead of regenerating blindly.

Pro ($10/mo, $8 annual) is the working creator's tier: 2,500 credits, v5.5 and the full model lineup, commercial rights, Voices, up to 3 Custom models, stem splitting, longer audio uploads, and a priority queue. Premier ($30/mo, $24 annual) is the production tier: 10,000 credits plus Suno Studio 2.0 - MIDI recording, wavetable synth, effects and plugins, automation, advanced stem separation, and unlimited 32-bit multitrack export. The commercial rights are identical between them; the difference is volume and tooling.

Credits are the single currency: a full song generation costs about 5 credits, so Free's 50/day is roughly 10 songs, Pro's 2,500/month roughly 500, and Premier's 10,000 roughly 2,000. Credits don't roll over - daily credits reset each day and monthly credits reset each billing cycle - while purchased top-ups persist with an active subscription. To stretch them: write specific prompts to cut regenerations, use the My Taste wand for better first takes, extend and edit good takes instead of regenerating from scratch, and lean on Studio 2.0's editing tools (which don't cost credits) rather than rolling the dice again.

The core promise: if you can describe a feeling, you can make a song. No instruments, music theory, or studio required - one sentence produces a finished track with vocals in under a minute, and each generation gives you two takes to choose from. From there you can extend, remix, edit stems, or publish to the community. It's the reason Suno is a top-10 music app and spans everyone from first-time creators to professional producers.

Covers reimagine an existing track in a completely different style while keeping its core melody and structure - turn your acoustic demo into synthwave, or your rap into bossa nova. You can cover your own Suno songs or audio you upload and hold rights to; it doesn't license you to cover copyrighted commercial songs. On Premier, Stem Cover takes it further by restyling individual stems inside Studio.

The iPhone app - "Suno - AI Songs & Music" by Suno, Inc. - and the Android app - "Suno - AI Music & Songs Maker", needing Android 5.0 or newer - are both free with in-app subscriptions. Create from a prompt, hum a melody, tap a beat, or record audio; browse the community; and manage your library. Subscriptions bought in the app work on the web too, and everything syncs to one account. Note that Studio 2.0 is desktop and tablet only - it needs at least a 768 px screen.

Because a genre is just words in a prompt, Suno makes style exploration nearly free: blend genres that rarely meet ("big band swing with synthwave production"), travel eras ("1970s Bollywood disco"), or push into the weird with the weirdness slider. The Explore feed, trending songs, and curated collections surface what the community is making, and Reuse Prompt lets you capture the vibe of a song you love in your own creation. My Taste then learns which styles you gravitate toward.

Beyond text, Suno builds songs from sound: hum a melody, tap out a beat, record a voice memo, or upload audio, and the model uses it as the seed - as a reference for vibe, a backing track to sing over, or raw material to remix. Free accounts can upload short clips (around a minute), while paid plans support much longer uploads. With v5.5, your voice itself becomes an input via Voices; and with Studio 2.0 you can upload MIDI files or play parts in live over Web MIDI.