LALAL.AI
AI stem splitting and vocal removal, online, on desktop or inside your DAW
Overview
LALAL.AI splits songs and videos into stems such as vocals, drums, bass, piano, guitars, synth, strings and wind, and cleans voice recordings of noise, echo and reverb. It runs on the web, desktop and mobile apps, as VST3 plugins and through an API, with a free preview tier and Lite and Pro subscriptions. It is operated by OmniSale GmbH in Switzerland.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 10 minutes in the Relaxed Queue
- 200 MB upload limit per file
- Result previews; no downloads
- Unlimited minutes in the Relaxed Queue
- 90 Fast Queue minutes a month
- 2 GB uploads, result downloads, batch processing
- 1 voice pack slot
- 250 Fast Queue minutes a month
- VST plugins and API access
- Local processing with Lyra in the desktop app and plugins
- 3 voice pack slots and early access
- Master: 750 Fast Queue minutes for $50
- Premium: 3,000 minutes for $190
- Enterprise: 5,000 minutes for $300
Use cases
What you can produce with LALAL.AI
- A clean acapella and matching instrumental extracted from any song or video
- A track split into stems such as drums, bass, piano, guitars, synth, strings and wind
- Lead and backing vocals separated, plus a backing-vocals-and-music mix
- Voice recordings cleaned of background music, noise, echo and reverb
- Stems separated inside a DAW with VST3 plugins that process locally
- A reusable clone of your own voice for voiceovers and ad reads
- Stem separation built into your own product through the LALAL.AI API
LALAL.AI remains one of the most complete stem splitters: ten stem types, a lead and backing vocal split, voice cleaning and dereverb, all on its Andromeda model, available in the browser, desktop and mobile apps, DAW plugins and an API. Pro adds on-device processing with the Lyra model, which spends no minutes, a real advantage for producers working with unreleased material.
The catch is metering. Minutes are charged per stem type, the free tier gives previews without downloads, and fast processing is capped each month. Worth paying for if you split audio regularly; occasional users should test on the free tier first.
What we found
LALAL.AI is a browser-based stem splitter run by OmniSale GmbH, a company registered in Risch-Rotkreuz, Switzerland. Its about page dates the service to 2020, when its first separation network, Rocknet, was trained on 20 TB of audio. The models have turned over repeatedly since: Cassiopeia, Phoenix, Orion and Perseus, and in 2025 Andromeda, which the vendor calls its sixth-generation engine and now the default for vocal and instrumental, voice and noise, lead and backing vocals, and echo and reverb removal. Earlier versions of this page named Phoenix as the current model, which is several generations out of date.
The stem menu is wider than most: vocals, drums, bass, piano, acoustic and electric guitar, synthesizer, strings and wind, plus a voice-and-noise split. Around it sit a Voice Cleaner, an Echo & Reverb Remover, a lead and backing vocal splitter, a Voice Changer and, since 2025, a Voice Cloner. There are desktop apps for Windows, macOS and Linux, iOS and Android apps, three VST3 plugins and a developer API. LALAL.AI says registered users rose from 3.8 million in 2024 to 6.79 million in 2025; that is its own figure.
The pricing model has changed, and older descriptions of it as pay-as-you-go with no subscription no longer fit. Individuals choose between a free Starter account, Lite at $9.99 a month and Pro at $19.99 a month, both cheaper when billed annually. Paid plans include unlimited processing in a slower Relaxed Queue plus a monthly allowance of Fast Queue minutes. Minutes are charged as file length times the number of stem types selected, so a four-minute song split three ways costs twelve. One-time top-up packs, from 750 minutes for $50, add fast processing.
The free tier is for judging quality, not finishing work: ten minutes with previews and no downloads. Pro is the plan for musicians who want local processing, because the Lyra model in the desktop app and plugins runs on their own machine without spending minutes.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including LALAL.AI's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with LALAL.AI unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to LALAL.AI directly →
Spotted something out of date? Suggest an update →
More in Video & Audio
