LALAL.AI — Online stem splitter that isolates vocals and instruments from any track
What it is
LALAL.AI extracts vocals, drums, bass and other stems from audio and video files using trained separation models, with per-minute pricing. It is a go-to for remixers, karaoke makers and DJs cleaning samples. Very dense mixes can leave artifacts, and the free preview is only short.
Editor's review
Long-form introduction by the BetterPicker editors · checked against the official site · Oct 18, 2026
LALAL.AI is a stem-separation studio that lives in a browser tab and treats audio surgery as a routine utility. Upload a track and the engine pulls out vocals, drums, bass, guitar, synth, string and wind instruments, or strips background music, mic rumble and echo from a voice recording. Around the splitter sits a full voice toolkit - cleaner, changer, cloner and a lead and backing splitter - with desktop and mobile apps so one subscription follows you across devices. The free Starter plan hands you ten relaxed processing minutes to judge the output before any money moves, and the paid Lite and Pro tiers lift the ceiling with unlimited relaxed minutes plus a monthly fast-queue allowance. Pricing is regional and shown at checkout, so the decision that actually matters is not the sticker but the queue: fast minutes are finite and reset every month, relaxed minutes are not.
What it does well
Separation breadth with honest metering: one upload can target eight stem types, and the pricing FAQ states the formula plainly - file length times the number of stem types selected - so a two-stem job costs half of a full separation and you can budget before uploading. Paid uploads reach 2GB per file, batch processing arrives on paid tiers, and the splitter handles everything from full mixes to voice memos.
A voice bench beyond splitting: Voice Cleaner removes backing music, plosives and mic rumble from talking recordings; Voice Changer and Voice Cloner reshape or rebuild a voice from samples; Echo and Reverb Remover dries out roomy takes. The same account balance works on the web app, the Windows, macOS and Linux desktop apps, and the iOS and Android apps, which makes it practical to start a job on a phone and finish it in a DAW.
Queue economics you can plan around: relaxed and fast queues produce identical quality - the only difference is priority - and paid plans include unlimited relaxed minutes plus 90 (Lite) or 250 (Pro) fast minutes a month. One-time top-ups, 750 minutes for $50 or 3,000 for $190, cover a heavy week without changing plans.
Who it's for
Remixers, karaoke makers, DJs and cover producers pulling acapellas and instrumentals from tracks they do not own the sessions for; podcast and interview editors cleaning noisy voice tracks; and musicians who want a VST plugin inside their DAW or an API in their product pipeline, both of which sit on the Pro tier. Teachers ripping practice backing tracks and video editors isolating dialogue also fit the metering model well.
Where it falls short
Fast minutes expire: unused fast-queue allowance resets at each billing period and subscriptions are non-refundable, so a quiet month is money left on the table and heavy months need top-ups rather than a bigger plan.
Dense mixes leave artifacts: busy arrangements can smear around cymbals and reverb tails, and the free Starter plan caps uploads at 200MB with only ten relaxed minutes - enough to judge, not enough to work.
The pro tools are gated: the VST plugin, API access and batch processing all require Pro, and extra voice pack slots top out at three, which matters for cloners juggling several personas.
Specs at a glance
Facts from the official site · not editorial opinion
| What it is | Browser and app stem-separation studio: vocal remover, stem splitter across eight instrument families, voice cleaner, voice changer, voice cloner, echo and reverb remover and a lead/backing vocal splitter |
|---|---|
| Price | Starter free with 10 relaxed minutes; Lite and Pro add unlimited relaxed minutes plus 90 / 250 fast minutes per month, billed annually - regional pricing shown at checkout |
| Platforms | Web plus desktop apps for Windows, macOS and Linux and mobile apps for iOS and Android, with one shared minute balance across all of them |
| Uploads | 200MB per file on Starter, 2GB on paid plans, with batch processing on paid tiers |
| Top-ups | One-time fast-minute top-ups: 750 minutes for $50, 3,000 for $190, 5,000 for $300 |
| Pro extras | VST plugin for DAWs and developer API access are Pro-plan features; up to 3 voice pack slots |
| Metering | Processing minutes = file length x number of stem types; fast minutes reset each billing period and do not roll over |
Frequently asked questions
▸Is LALAL.AI free to use?
Starter is always free with ten relaxed processing minutes and 200MB uploads. Paid Lite and Pro plans add unlimited relaxed minutes, monthly fast minutes, larger uploads and batch processing.
▸What is the difference between the fast and relaxed queues?
Output quality is identical. Fast Queue processes with priority; Relaxed Queue waits for server capacity. Paid plans start each month in Fast Queue until the allowance runs out, then switch automatically.
▸Do unused minutes roll over?
No. Fast-queue minutes reset at the start of each billing period. Relaxed minutes are unlimited on paid plans, so there is nothing to roll over.
▸Can I use one subscription on several devices?
Yes. The subscription is tied to your account and works across the web, desktop and mobile apps with a shared minute balance.
Video reviews
3 review videos aggregated · praise and criticism included alike · click through to the original video
Channels that covered it
Channels are aggregated as sources only — we don’t rate creators
Related tools
Where to go next
External links open in a new tab; external content is independent of this site.
Link down? Every object page is re-checked monthly.