Explainer
What is an AI studio?
An AI studio is a production workspace with generative AI built in. In music, it is a multitrack studio, much like a DAW, where AI can create parts (a beat, a vocal, a whole song, a sound effect) straight onto the timeline next to the parts you record or play, and then help you mix, master and export a finished track.
What “AI studio” means
Studio has always meant the place where a recording gets made: the instruments, the microphones, the desk that mixes them, and the people who know how to use it all. A DAW (digital audio workstation) moved that studio into software. An AI studio adds a new kind of collaborator to the software: models that can write and play a part when you describe it.
That changes where material comes from, not how a song is finished. You still arrange sections, balance levels and decide what stays. The difference is that a missing drum part, a backing vocal or a whole first draft can come from a prompt or a reference instead of a session.
The phrase is used elsewhere too. Developers say “AI studio” for web consoles where you try out and tune AI models, and designers use it for AI image and video tools. On this page it means an AI music studio.
AI studio, DAW or AI generator?
Three kinds of music tool, and what each one is like to work in.
| Question | Traditional DAW | One-shot AI generator | AI studio |
|---|---|---|---|
| Where the music comes from | What you record, play, program or import | A prompt; the AI makes the whole track | Both: your own takes and parts, plus AI-generated ones on the same timeline |
| Changing one part | Edit it by hand, note by note or clip by clip | Usually means generating the whole song again | Regenerate, extend or replace just that section, stem or clip |
| What you get | A multitrack project you finish yourself | A finished audio file | A multitrack project, finished or not |
| Mixing and mastering | Effects and your ears | Baked into the file | Built-in effects, plus assisted mixing and mastering you can adjust |
| Skills to start | Production know-how | None: type a prompt | None to start, as deep as you want to go |
What you can do in Soundverse AI Studio
Soundverse AI Studio is a multitrack studio in your browser for audio, MIDI, video and image tracks, with AI generation on the timeline.
Record and play
Record vocals and instruments from your microphone, play eight built-in instruments, and edit their notes in a piano roll. Synth One, Bass One, Drum Kit One and Sampler One also take a MIDI controller.
Generate on the timeline
Right-click an empty part of a track to generate an instrumental, a song, vocals, sound effects or dialogue. Right-click a clip to extend it, remix it or split it into stems; remixing and stems need the Creator plan. Each result lands as a clip you can trim, move and mix.
Arrange and edit
Split, trim and move clips across audio, MIDI, video and image tracks. Every clip that arrives has its tempo and key detected, and Follow tempo stretches a clip to the project's tempo while keeping its key.
Mix and master
Use 28 built-in effects, from EQ and compression to reverb and delay, and automate volume, pan and effects over time. Mix One places each track by genre template; Mastering One listens to your mix and sets itself toward a genre and loudness target. Both run in your browser at no token cost.
Score a video
Video and image lanes sit on the same timeline as the music. Describe and Generate Audio reads a video and adds the music, sound effects and dialogue its scenes call for; Video to SFX adds effects timed to the action. Both need the Creator plan.
Work together, live
Invite anyone by email or share a link. Everyone sees live cursors and edits, and can comment on a track at a bar, chat, and start a call with up to three other people who are online.
Ask Agent One
Open Agent One from the top bar and ask for a part, a new track or an arrangement change. With its Studio switch on, it edits the project live, adding tracks, notes and clips you can keep editing.
What’s inside Soundverse AI Studio
The instruments, the effects, Mix One and Mastering One are built into the studio and run in your browser, on every plan, with no tokens. The AI tools on the timeline are models that run on servers, and they use tokens.
All 8 instruments
Each instrument takes one track and plays into your export. Synth One, Bass One, Drum Kit One and Sampler One also take a MIDI controller.
Synth One
A wavetable synth for pads, leads, keys and basses, with 100 factory presets.
Bass One
A synth bass for sub, 808, acid and reese lines, with 41 presets, glide and unison.
Drum Kit One
A 16-pad drum synth with 10 kits, from trap and house to techno and cinematic. You program its beats in the piano roll.
Sampler One
Plays any sample chromatically, as a one-shot, or cut into 16 slices, from WAV, MP3, OGG, FLAC or AIFF files up to 100 MB.
Guitar One
A MIDI guitar on Synth One's engine: 6, 7 or 8 strings, many tunings, a 24-fret fretboard and 52 presets.
Chord One
Key-aware chords and progressions in 12 keys, 8 modes and 7 genres, which you drag onto any instrument track. Its suggestions come from music-theory rules.
Visual One
Play chords with your hands through your webcam, or with the mouse; opening your mouth sweeps a filter. The hand and face tracking runs in your browser.
VocalMIDI One
Sing, hum or beatbox and it turns your voice into MIDI as you perform. Your microphone's sound never leaves the browser.
Watch it: 4:42 Bass One synth in the walkthrough
All 28 effects
Add effects from a track's ＋ Add effect. New tracks start with EQ Eight, Reverb • Space and Delay • Ping-Pong, the last two switched off, and the master bus starts with Mastering One, off. One click on an effect-chain preset (34 of them, for vocals, guitar, drums, bass, and piano and keys) sets up a whole chain.
Dynamics
- CompressorEvens out level so quiet and loud parts sit together.
- LimiterA ceiling that catches peaks so nothing clips.
- Sidechain OneTempo-synced ducking that makes a track pump.
- De-Esser OneTames harsh “s” sounds in a vocal.
EQ, filter and level
Space and stereo
- Reverb • SpaceRoom and hall reverb, as a parallel send.
- Reverb OneHall, room, plate, chamber or ambience, with freeze and width.
- Delay • Ping-PongTempo-synced echoes, as a parallel send.
- Ping Pong DelayStereo echoes that bounce from side to side.
- Delay OneTempo-synced delay in single, dual, ratio, ping-pong or quad styles.
- WidenerOpens up the stereo image.
- Auto PanSweeps the sound across the stereo field in time.
Character and drive
- Drive • TubeOverdrive, from a warm break-up to full distortion.
- SaturatorWarm harmonic saturation.
- Bitcrush • Lo-fiFewer bits for a crunchy, vintage-digital sound.
- Vinyl OneDust, scratch, warp, wear and rumble, set by era.
- Cassette OneTape wow and flutter, saturation, hiss and head bump.
- Guitar One (amp)Amp, cabinet and pedals with genre voicings.
Creative and vocal
- Chorus • EnsembleThickens and widens with detuned doubling.
- Phaser • 4-StageSweeping notches for a swirling texture.
- Stutter OneBeat-synced stutter, glitch, gate and tape-stop.
- VocalDuality OneA vocal doubler with spread and detune.
- VocalTune OnePitch correction you hear as you play; exports stay dry.
Mastering and metering
- Mastering OneThe master bus's mastering chain, with an automatic ✦ Master (below).
- Analysis OneTen meters: spectrum, spectrogram, goniometer, correlation, waveform, loudness (LUFS), peak and RMS, stereo bands, resonance and waterfall. It changes no sound.
Watch it: 5:13 Effects: reverb and EQ in the walkthrough
Watch it: 4:52 Automation: filter cutoff and volume in the walkthrough
Automatic mixing and mastering
Mix One and Mastering One are automatic, not generative AI. Both measure your audio in the browser and apply rules and formulas to it, so they cost no tokens on any plan, and you can change anything they set.
Mix One: automatic mixing
Mix One is the studio's mixing stage: open it from the ☰ menu, under Workspace. It draws every track as a block on a chart of frequency against stereo position. Set each track's type (drums, bass, vocals and so on), pick a genre template, and Auto-Place seats each typed track in its place for that genre, then pans clashing parts apart and carves room between them. Tracks without a type stay where you put them.
It names the tracks that mask each other, and Fix tone applies the corrections it found: a high-pass, cuts for mud, boxiness, harshness and honk, and de-essing. Nothing changes until you switch it on, and you can keep three versions (A, B and C) to compare. What you hear is what you export.
It places and carves; it doesn't balance your levels, which stay with your faders and its trim.
18 genre templates
- Pop / Universal
- Hip-Hop / Trap
- R&B / Neo-Soul
- House / Melodic EDM
- Techno / Minimal
- Drum & Bass
- Indie / Alt Rock
- Metal / Hard Rock
- Acoustic / SSW
- Jazz / Live
- Filmi / Bollywood
- Punjabi / Desi Pop
- Indian Devotional
- Afrobeats / Amapiano
- Reggaeton / Latin
- Lo-fi / Chill
- Cinematic / Score
- Ambient / Meditation
Watch it: 6:36 Mix: the mixer, then Mix One in the walkthrough
Mastering One: automatic mastering
Mastering One sits on every project's master bus, off until you switch it on (Mixer, then the master strip's FX). From your Library, a song's ⋯ menu → Master opens the song in a new project with it on. It is a chain of tone EQ, vocal presence, a stabilizer, two-band dynamics, a stereo imager, an exciter and a maximizer, run by six controls.
Pick a genre and a loudness target, play your song, and press ✦ Master: it listens to 2 seconds of the mix and sets five of the six controls (loudness, tone, width, dynamics and intensity) toward that genre and target, leaving vocal balance to you. Then fine-tune, or start from a preset. A peak ceiling and 16-bit dither finish the chain.
7 genres
- Pop
- Hip-Hop
- EDM / House
- Rock
- R&B / Soul
- Acoustic
- Cinematic
8 loudness targets (LUFS)
- Club / DJ −7
- EDM −8
- Hip-hop loud −9
- Rock −10
- Loud −12
- Streaming −14
- Apple Music −16
- Broadcast EBU R128 −23
6 presets
- Streaming −14
- Apple Music −16
- Hip-Hop Loud −9
- Club / DJ −7
- Broadcast −23
- Transparent
Watch it: 7:10 Mastering One in the walkthrough
Measure it, and see it
Analysis One
A metering plugin for any track or the master: spectrum, spectrogram, goniometer, correlation, waveform, loudness in LUFS, peak and RMS, stereo bands, resonance and a waterfall. It measures and changes no sound.
Visualizer
Audio-reactive visuals for your song, from the ☰ menu under Workspace: 10 scenes (Cassette Deck, Liquid Chrome, Ring, Outrun, Spectrum, Warp, Ink Water, Piano Roll, Symmetry and Spectrogram), with layers, effects and a node canvas for the visual rig, rendered to video at 16:9, 9:16, 1:1 or 4:5.
AI on the timeline
In the studio, AI means models that make new audio, images and video. Choose a tool from a right-click and press Generate in Studio: the result lands on the timeline as a new clip or track, never over one of yours. These tools use tokens at each tool's price. Most run on Free; the ones marked Creator need the Creator plan or above.
Make something new
Right-click an empty part of a track, or the video or image lane.
- AI MusicAn instrumental at the playhead.
- AI SongA song with vocals; it can write the lyrics first.
- VocalsA sung part over what is already there.
- SpeechA spoken line or narration.
- Sound EffectA sound effect of the length you set.
- AI ImageAn image on the image lane.
- AI VideoCreatorA video clip on the video lane.
Work from a clip
Right-click an audio clip, or select part of one.
- ExtendContinues a clip, as a song or as music.
- RemixCreatorThe clip restyled, on a new track.
- Similar MusicCreatorA new instrumental in the clip's style.
- Similar SongCreatorA new song in the clip's style.
- Similar VocalsCreatorVocals in the tone of a reference voice.
- Melody to SongA full song from a hummed or played melody.
- InpaintingCreatorRewrites a selected part and keeps the rest.
- Voice SwapThe vocal, sung again in another voice.
- Voice to InstrumentA hummed or sung line, played on an instrument.
- Stem GeneratorCreatorOne new stem that fits the clip, from 12 kinds.
- Separate StemsCreatorThe clip split into 2, 6 or 14+ stems, each on its own track.
- Audio to MIDIA MIDI track for each instrument it hears, played on the studio's instruments.
Work from a video
Right-click a video clip.
- Video DescriberCreatorDescribes the clip's picture, sound effects, music and dialogue.
- Video to SFXCreatorSound effects timed to the action, on new tracks.
Three more, straight from a clip's menu
Section Analysis writes a song's tempo, sections and beats onto its clip. Describe and Generate Audio scores a video with music, sound effects and dialogue on new tracks (Creator). Generate Music for Video opens AI Music with the video's own music description, to span the clip.
Agent One in the studio
Open Agent One from the studio's top bar and ask in words. With its Studio switch on (it is, in each new project), it works in the project beside you: it writes MIDI parts on a fitting instrument (drums on Drum Kit One, bass on Bass One, chords, keys, pads and leads on Synth One, guitar on Guitar One), moves, trims and splits clips, adds tracks, sets effects and EQ, draws automation, arranges your images and video on their lanes, and generates audio straight onto the timeline. It listens to a clip's key, tempo and sections before it writes, and shows up as a collaborator with its own cursor. It can't change the project's tempo or run Mix One. Its edits cost nothing; the Agent One run itself uses tokens.
Recording, editing, collaboration and export
The rest of the studio, around the timeline: what you record and play, how you arrange it, who works on it with you, and what comes out.
Recording
Arm any number of audio and MIDI tracks and record from the playhead. Each audio track takes its own input, with live monitoring. There's no count-in or punch-in yet; VocalMIDI One has a count-in of its own.
MIDI
Play notes from the on-screen keys, your computer keyboard or, on Synth One, Bass One, Drum Kit One and Sampler One, a MIDI controller, and edit them in the piano roll with a velocity lane. Drag a .mid file from your Library onto a MIDI track to bring it in, and save any MIDI clip as a .mid file.
Arranging and tempo
Copy, cut, paste, duplicate, split and mute clips. Every clip that arrives has its tempo and key detected in your browser; Follow tempo stretches it to the project's tempo without changing its key, and pitch shifts up to 12 semitones either way. Nine layouts, from Songwriting to Mixing and Music Video, open the panels each job needs.
Automation
Draw automation lanes for volume, pan and effect settings on any track; the curves play back and render into your export.
Stems
Separate Stems splits a clip into 2, 6 or 14+ stems, each on its own track and lined up with the source, and Stem Generator adds one new stem that fits; both need Creator. Exporting your own tracks as stems is separate: a ZIP with a file per track, from Creator.
Real-time collaboration
The owner and Admins invite anyone by email, as an Editor, Viewer or Admin (an invite to a new address lasts 7 days), or share a link that can view or edit. Everyone sees named cursors and each other's edits as they happen, comments on a track at a bar, chats, drops shared files onto the timeline, and can start an audio or video call with up to three other people who are online. Collaborators need only a free account; a project's track and video-lane limits follow its owner's plan, and AI tools and export follow each person's own.
Video and image lanes
Video and image lanes sit on the same timeline as the music: generate or import video and images, arrange them, and export the picture with the sound. Free projects get one video lane; paid plans have no lane limit.
Export
Export runs in your browser: audio as WAV, MP3, FLAC, AAC or OGG at 44.1, 48 or 96 kHz, with an optional loudness target (Streaming −14, Club −9 or Broadcast −23 LUFS); stems as a ZIP of WAV, FLAC or MP3 files, with or without each track's effects; and video as MP4, WebM or MOV at 24, 30 or 60 fps, from 720p up to 4K.
Watch a song made start to finish
One song from a piano loop to a mixed and mastered track, then a music video for it, all in Soundverse AI Studio. Jump to any part.
- Start: a new project, a name and a tempo
- Upload a piano loop
- Extend Music
- Tidy: delete, rename, split
- Bass with the AI Stem Generator
- Strings: a new stem
- Festive pop drums with sleigh bells
- Vocals: lyrics, a voice, a song
- Inpainting: rewrite one line
- Audio to MIDI: the bass as notes
- Bass One synth
- Automation: filter cutoff and volume
- Effects: reverb and EQ
- Clean-up: cut stray sounds
- Detail: drums, strings and vocal
- Mix: the mixer, then Mix One
- Mastering One
- Final balance pass
- The finished song: Santa's Sleigh Ride
- AI Music video generator with Veo 3.1
- The cast: reference images from the Library
- Scenes: one shot per section
- Arrange every scene on its bars
- Export: 1080p MP4
Who an AI studio is for
Songwriters
Hear a lyric or a melody as a full arrangement, then rewrite one line or one section without starting over.
Producers and beatmakers
Generate the missing part, split a sample into stems, program MIDI and mix, all in one project.
Singers
Record over a generated beat, layer harmonies, and export the song or its stems.
Video creators
Put the picture on the timeline and build its music, sound effects and dialogue to the frame.
Bands and remote collaborators
Work in one shared project instead of emailing stems and voice notes back and forth.
How to start
Four steps from a blank project to an exported song, in Soundverse AI Studio.
Open a project
Sign up free and click Studio in the sidebar. It opens your latest project, or a new one with a Quick start panel for importing, uploading or recording your first file.
Add tracks, sounds and takes
Click Add Track for an audio or MIDI track, drag files in, or arm a track and press Record. Load an instrument from Instruments in the sidebar.
Generate on the timeline
Right-click an empty part of a track and choose what to generate, or ask Agent One from the top bar. The result lands on the timeline, ready to edit.
Mix, master and export
Add effects, balance levels in the Mixer, place tracks with Mix One, master with Mastering One, then click Export for audio, stems or video.
What it costs
Soundverse AI Studio is included on every plan, Free too. The studio itself (instruments, effects, Mix One, Mastering One, recording, editing and export) uses no tokens; AI generation and Agent One do, and some AI tools need the Creator plan or above.
Your plan sets how many projects you keep, how many tracks each holds and what you can export:
| AI Studio | Free | Creator | Pro | Max |
|---|---|---|---|---|
| AI Studio projects | 3 | 8 | 20 | 30 per Max level, up to 150 |
| Tracks per project | 3 | 8 | 20 | 30 per Max level, up to 150 |
| Video lanes per project | 1 | No plan limit | No plan limit | No plan limit |
| Audio export | MP3 | MP3, WAV, FLAC, AAC, OGG | MP3, WAV, FLAC, AAC, OGG | MP3, WAV, FLAC, AAC, OGG |
| Stems export | Not included | Included | Included | Included |
| Video export | Up to 720p | Up to 1080p | Up to 4K | Up to 4K |
Prices, tokens and each plan’s license are on the Pricing page, for every plan and in your currency.
What it can’t do yet
- Its instruments and effects are built in, so it does not load third-party VST or AU plugins.
- Recording has no count-in or punch-in yet, and no take comping.
- Export in the browser is blocked for projects longer than 15 minutes or with more than 24 tracks.
- Video export needs the browser's own video and audio encoders, which current desktop Chrome and Edge have.
- VocalTune One corrects pitch as you listen, but exports are dry.
- You can automate volume, pan and effect settings, but not instrument knobs.
- Agent One can't change the project's tempo or run Mix One.
- The Viewer role and view-only links don't lock the timeline, so invite only people you'd trust to edit it.
- Calls take up to three other people who are online, and have no screen sharing.
- Project and track counts follow your plan; a full project needs fewer tracks or a bigger plan before you add more.
Frequently asked questions
What is an AI studio?
An AI studio is a production workspace with generative AI built in. In music, it is a multitrack studio like a DAW where AI can create parts, from a beat or a vocal to a whole song, directly on the timeline next to your own recordings, and help you mix, master and export.
Is an AI studio the same as a DAW?
It is a kind of DAW. A traditional DAW records, edits and mixes the material you bring to it; an AI studio does the same and can also generate material onto the timeline, so a missing part or a first draft can come from a prompt or a reference.
How is an AI studio different from an AI song generator?
An AI song generator gives you a finished audio file, and changing one part usually means generating the song again. An AI studio puts what the AI makes onto tracks, so you can keep the parts you like, regenerate one section or stem, record your own takes over it and mix the result.
Do I need production skills to use an AI studio?
No. You can start from a prompt or let Agent One build a first version, then learn the studio tools as you go: trimming clips, adding effects, balancing tracks. Experienced producers can work hands-on from the first click.
Is Soundverse AI Studio free?
Yes. AI Studio is included on every Soundverse plan, including Free, which keeps 3 projects with up to 3 tracks each and exports MP3 audio and 720p video. The studio's own tools, from the instruments and effects to Mix One and Mastering One, use no tokens; AI generation uses tokens, and some AI tools need Creator or above.
Which instruments and effects does Soundverse AI Studio include?
Eight instruments: Synth One, Bass One, Drum Kit One, Sampler One, Guitar One, Chord One, Visual One and VocalMIDI One. And 28 effects, from Compressor, EQ Eight and Reverb One to Vinyl One, Stutter One, VocalTune One, Analysis One and Mastering One, plus Mix One for mixing. All of them run in your browser on every plan, at no token cost.
Does Soundverse AI Studio mix and master automatically?
Yes. Mix One places each typed track by one of 18 genre templates, carves clashing parts apart and can fix tone problems it finds. Mastering One's ✦ Master listens to 2 seconds of your mix and sets itself toward a genre and a loudness target. Both are automatic rather than generative AI: they run in your browser, use no tokens, and you can change everything they set.
Can I use a MIDI controller?
Yes, with Synth One, Bass One, Drum Kit One and Sampler One. The other instruments play differently: Guitar One from your computer keyboard, Chord One by dragging chords onto a track, Visual One with your hands through the webcam or the mouse, and VocalMIDI One with your voice.
Can I record my own vocals and instruments?
Yes. Arm a track and press Record to capture your microphone, or play the built-in instruments; four of them also take a MIDI controller. Your takes sit on the same timeline as anything the AI generates.
Can I work on a project with other people?
Yes. The owner or an Admin invites anyone by email as an Editor, Viewer or Admin, or shares a link. You edit the same project live, with each other's cursors, comments on tracks, a chat and calls with up to three other people who are online. Collaborators need only a free account.
Does Soundverse AI Studio run in the browser?
Yes. It runs in your web browser with nothing to install, and projects are saved as you work, so you can open them again from any computer.
Can I use what I make commercially?
It depends on your plan. Music made on the Free plan is for non-commercial use only. What each paid plan allows differs; the Terms of Service and the Pricing page set it out. You need the rights to anything you upload or record into a project.
Try an AI studio on your next song
Soundverse AI Studio is free to start and runs in your browser. Bring a loop, a lyric or nothing at all.