Description
LALAL.AI
Overview
LALAL.AI is a next-generation AI-powered audio source separation service, recognized as the #1 stem splitter worldwide. It allows users to extract vocals, instrumentals, drums, bass, piano, guitars, synthesizers and more from any song or video without quality loss.
Founded in 2020 and operated by Omnisale GmbH with legal headquarters in HolzhƤusern, Switzerland and development origins in Nizhny Novgorod, Russia. Co-founded by Dmitry Orlov.
In 2025, LALAL.AI reached 6.79 million registered users (up from 3.8 million in 2024), marking the strongest year-on-year growth in the company's history.
Products and Tools
Stem Splitter (Core Product)
| Stem | Description |
|---|---|
| Vocal & Instrumental | Classic voice/music separation |
| Voice & Noise | Voice/background noise cleaning |
| Drums | Drum extraction |
| Bass | Bass extraction |
| Piano | Piano extraction |
| Electric Guitar | Electric guitar |
| Acoustic Guitar | Acoustic guitar |
| Synthesizer | Synthesizers |
| Strings | String instruments |
| Wind | Wind instruments |
Additional Tools
| Tool | Description |
|---|---|
| Voice Cleaner | Background noise, rumble, plosives removal |
| Voice Changer | Voice transformation with artist clones |
| Voice Cloner | Custom voice replica creation |
| Echo & Reverb Remover | Echo and reverb elimination |
| Lead & Back Vocal Splitter | Lead and backing vocal separation |
Neural Networks (Generations)
Technology Evolution
| Network | Year | Features |
|---|---|---|
| Rocknet | 2020 | First network, 20TB training data |
| Cassiopeia | 2021 | Fewer artifacts, superior quality |
| Phoenix | 2022 | Record speed, mask-based method |
| Orion | 2023 | Direct synthesis, 70% fewer artifacts |
| Perseus | 2024 | Transformer architecture (like ChatGPT), 15% better quality |
| Andromeda | 2025 | 4x more data, 40% faster, 10% better SDR |
Andromeda (Latest Generation)
- Advanced transformer architecture
- Processing up to 40% faster than Perseus
- 10% better SDR (Signal-to-Distortion Ratio)
- Eliminates trade-off between detail and bleeding
- Default in Stem Splitter, Voice Cleaner, Echo Remover
Supported Formats
Input (Upload)
| Type | Formats |
|---|---|
| Audio | MP3, OGG, WAV, FLAC, AIFF, AAC, M4A |
| Video | AVI, MP4, MKV, MOV, M4V |
Output (Download)
- Same format as original file
- Options: MP3, WAV, FLAC, OGG, AAC, AIFF
Available Platforms
| Platform | Availability |
|---|---|
| Web | lalal.ai (browser) |
| Windows | Desktop app |
| macOS | Desktop app |
| Linux | Desktop app |
| iOS | Mobile app |
| Android | Mobile app |
| API | For enterprise integration |
| VST Plugin | In development (2026) |
Pricing Plans
Minute Packs (One-Time Purchase)
| Plan | Price | Minutes |
|---|---|---|
| Starter | Free | 10 min (preview only) |
| Lite | $20 | 90 min |
| Plus | $27 | 300 min + 30 bonus |
| Master | $50 | 750 min |
| Premium | $190 | 3,000 min |
| Enterprise | $300 | 5,000 min |
Subscriptions (Monthly/Annual)
- Modes: Relaxed (queue) and Fast (immediate)
- Fast minutes limited per plan
- Relaxed unlimited for subscribers
- 50% discounts on special promotions
Enterprise
- Custom pricing
- Dedicated API
- Premium support
- Custom integration
Technical Features
Processing Modes
| Mode | Description |
|---|---|
| Clear Cut | Minimizes bleeding between stems |
| Deep Extraction | Captures finer details |
| Enhanced Processing | Additional neural network for better extraction |
| De-Echo | Pre-processing echo removal |
Noise Canceling Levels
- Mild - Soft reduction
- Normal - Standard reduction
- Aggressive - Intense reduction
Use Cases
For Musicians
- Create karaoke/instrumental tracks
- Extract stems for remixes and mashups
- Isolate parts for practice
- Sample specific elements
For Producers/DJs
- Prepare tracks for sets
- Create edits and bootlegs
- Isolate acapellas
- Remastering
For Content Creators
- Remove background music from videos
- Clean podcast audio
- Isolate dialogue
- Create voiceovers
For Education
- Analyze musical compositions
- Study vocal techniques
- Educational material
- Instrumental practice
Company Information
| Data | Value |
|---|---|
| Legal Name | Omnisale GmbH |
| Founded | 2020 |
| Legal HQ | HolzhƤusernstrasse 16, CH-6343 HolzhƤusern, Switzerland |
| Development | Nizhny Novgorod, Russia |
| Co-founder | Dmitry Orlov |
| Users (2025) | 6.79 million |
| Users (2024) | 3.8 million |
| Status | Active, growing |
| Funding | Bootstrapped (no external investment) |
Development Timeline
| Year | Milestones |
|---|---|
| 2020 | Founded, Rocknet, 2-stem splitter launch |
| 2021 | Cassiopeia, world's first 8-stem splitter, API, 8 languages |
| 2022 | Phoenix, Voice Cleaner, Desktop app |
| 2023 | Orion, iOS/Android apps |
| 2024 | Enhanced Processing, Voice Changer, Echo Remover, Lead/Back Splitter, Perseus |
| 2025 | Voice Cloner, Andromeda, 6.79M users |
| 2026 | VST Plugin (announced), Andromeda multi-stem |
Competitors
| Competitor | Comparison |
|---|---|
| Moises.ai | Similar functionality, powered by Deezer AI |
| AudioShake | #1 in Meta benchmarks, enterprise-focused |
| iZotope RX | Professional paid software |
| Spleeter | Open-source by Deezer, free |
| Ultimate Vocal Remover | Open-source, free, customizable |
| Demucs | Open-source by Meta, free |
LALAL.AI's competitive advantage: Combination of professional quality, ease of use, multiple platforms and accessible pricing.
Supported Languages
- English
- ē®ä½äøę (Chinese Simplified)
- ē¹é«äøę (Chinese Traditional)
- FranƧais
- Deutsch
- Italiano
- ę„ę¬čŖ
- ķźµģ“
- EspaƱol
- Nederlands
- PortuguĆŖs
- PortuguĆŖs Brasileiro
PROS ā
- 10 stems - Most in the market
- Professional quality - Latest generation Andromeda engine
- Multi-platform - Web, desktop, mobile, API
- Free preview - Listen before processing
- No subscription required - One-time purchase packs
- Multiple tools - Voice Cleaner, Changer, Cloner, etc.
- Diverse formats - Audio and video supported
- Batch processing - Multiple simultaneous files
- 6+ neural networks - Constant evolution
- 12 languages - Multilingual interface
- Desktop apps - Windows, Mac, Linux
- Enterprise API - B2B integration
CONS ā
- Limited free tier - Preview only, no downloads
- Credits don't accumulate - Fast minutes don't roll over
- Occasional artifacts - On complex mixes
- Per-minute pricing - Adds up with heavy use
- No VST yet - Plugin in development
- No rollover - Unused minutes are lost
- Variable quality - Depends on source audio
Key Features
Stem splitter 10 stems vocals instrumentals
AI vocal remover voice elimination
Andromeda neural network transformer
Voice Cleaner background noise removal
Voice Changer voice transformation
Voice Cloner custom voice cloning
Echo Reverb Remover elimination
Lead Back Vocal Splitter main vocals
Multi-platform web desktop mobile
Enterprise API B2B integration
Batch processing multiple files
Free preview before processing
Drums bass piano guitar extraction
Enhanced Processing Clear Cut Deep
Noise Canceling Levels mild normal aggressive
Formats MP3 WAV FLAC video MP4
Desktop apps Windows Mac Linux
Mobile apps iOS Android
12 languages multilingual interface
No subscription required one-time purchase
Use Cases
Karaoke instrumental tracks songs
Remix mashup music production
DJ sets track preparation
Acapella isolated vocals extraction
Sampling specific musical elements
Music practice individual stems
Music education composition analysis
Podcast audio voice cleaning
Video editing background music
Voiceover content creation
Remastering old audio improvement
Cover songs backing tracks
Music transcription learning
Sound design effects elements
Live performance preparation
Content creation social media
Film dubbing isolated dialogues
Gaming audio adaptation
Streaming legal music
Vocal training technique study
Information
Company
Omnisale GmbH
Website
www.lalal.ai