Descript
Text-based video and podcast editor with AI transcription, voice cloning, and filler-word removal
AI stem separation that splits any track into vocals, instrumental, drums, bass, and more
LALAL.AI splits a song into vocals, instrumental, and up to 10 stems using its sixth-generation Andromeda engine, plus voice cleaning and noise removal. The free Starter tier gives 10 minutes; paid subscriptions start around $9/month, and one-time minute packs run from $50 for 750 minutes to $300 for 5,000. Best for producers, remixers, karaoke makers, and podcast editors.
LALAL.AI is a web service that uses AI to separate a finished audio or video recording into its component stems. It began as a vocal remover and now splits a track into as many as ten parts — vocals, instrumental, drums, bass, acoustic and electric guitar, piano, synthesizer, strings, and wind — powered by a sixth-generation neural engine the company calls Andromeda. Alongside separation, it offers a Voice Cleaner that strips background noise, echo, and reverb, plus lead-versus-backing vocal splitting.
The service is metered in processing minutes rather than a flat subscription for unlimited use. Crucially, minutes are charged as the length of a file multiplied by the number of stems you extract, so a five-minute song split into three stems consumes fifteen minutes. Subscriptions provide a monthly pool of priority Fast Queue minutes plus unlimited slower Relaxed Queue processing, while one-time packs sell larger blocks of Fast Queue minutes with no recurring commitment. A public API and a VST plugin on the Pro plan let producers fold separation directly into their existing tools and digital audio workstations.
LALAL.AI suits anyone who needs to pull apart a mixed recording without owning the original session files. The free 10 minutes is enough to test quality on a single track, but real projects quickly push users toward a subscription or a minute pack.
Starting price: $9/mo · Free tier: yes · Model: freemium
Price history tracked from June 2026
| Plan | Price | Includes |
|---|---|---|
| Starter | Free | 10 minutes in the Relaxed Queue · 200 MB per-file upload limit · Result previews; full downloads restricted · No batch processing |
| Lite | $9/mo | Unlimited Relaxed Queue minutes · 90 Fast Queue minutes per month · 2 GB per-file upload · Batch processing · Cheaper billed annually |
| Pro | $19/mo | Unlimited Relaxed Queue minutes · 250 Fast Queue minutes per month · API access and VST plugin for DAWs · Batch processing, 2 GB per-file upload · Cheaper billed annually |
| Master (one-time pack) | $50 | 750 Fast Queue minutes · No subscription required |
| Premium (one-time pack) | $190 | 3,000 Fast Queue minutes |
| Enterprise (one-time pack) | $300 | 5,000 Fast Queue minutes |
| Pros | Cons |
|---|---|
| Separation quality is strong for a one-click web tool, with up to 10 distinct stem types | Minutes are consumed by file length times the number of stems, so a 5-minute song into 3 stems uses 15 minutes — heavy jobs drain packs fast |
| Works across audio and video files with a generous format list | Fast Queue minutes are monthly on subscriptions and do not roll over |
| One-time minute packs let occasional users avoid a subscription entirely | Free tier is only 10 minutes with capped uploads and restricted downloads, so any real use requires paying, and dense mixes can still bleed between stems |
| Free tier and result previews let you test quality before paying |
Text-based video and podcast editor with AI transcription, voice cloning, and filler-word removal
AI music generator that turns text prompts into full songs with vocals
AI voice platform for text-to-speech, voice cloning, and conversational agents in 70+ languages
AI music generator that creates full songs with vocals from text prompts
LALAL.AI has a free Starter tier with 10 minutes of processing in the Relaxed Queue and a 200 MB per-file limit. You can preview results, but full downloads and batch processing require a paid subscription or a one-time minute pack.
Subscriptions start around $9 per month for Lite, which adds 90 Fast Queue minutes monthly, and about $19 per month for Pro with 250 minutes plus API access. One-time packs run $50 for 750 minutes, $190 for 3,000, and $300 for 5,000.
Minutes are calculated as the length of your file multiplied by the number of stem types you select. Separating a five-minute song into three stems therefore uses fifteen minutes, so multi-stem jobs consume your allowance quickly.
It can split a track into vocals and instrumental, plus drums, bass, acoustic and electric guitar, piano, synthesizer, strings, and wind instruments. It also separates lead from backing vocals and can clean noise, echo, and reverb.
Yes. API access is included with the Pro plan, which also adds a VST plugin so you can run separations directly inside a digital audio workstation. The API lets developers integrate stem separation into their own apps and workflows.
Yes. It accepts video formats including MP4, MKV, MOV, AVI, and M4V in addition to audio formats like MP3, WAV, FLAC, and AAC, extracting and separating the audio track from the file you upload.