The new release improves Andromeda's capabilities to cover Drums and Bass and delivers up to 40% faster processing.
ZUG, CH / ACCESS Newswire / April 3, 2026 / LALAL.AI, a leading AI-powered audio separation platform, announces the release of the new version of Andromeda, a major update to its flagship neural network. The new release extends Andromeda's stem separation capabilities to include Drums and Bass - two of the platform's most requested stem types - while delivering significant improvements in separation quality and processing speed across all supported stems.
The update arrives as LALAL.AI reaches a series of notable industry milestones, cementing its position as the benchmark for professional-grade audio separation.
What's New in Andromeda
Drums rank among the top three most frequently separated stems on the LALAL.AI platform, with more than 5 million splits in 2025 alone. With the new Andromeda release, users now have access to a significantly improved drum extraction model within LALAL.AI's most advanced neural network, alongside the vocal, instrumental, and bass stems already supported.
Key improvements in Andromeda include:
Drums and Bass separation - added to the Andromeda model for the first time, enabling the full rhythm section to be isolated within a single neural network alongside Vocals and Instrumental.
Fuller bass reproduction - broader frequency range capture preserves harmonics, texture, and upper-register detail that frequently gets lost in separation.
Up to 40% faster processing - compared to the prior-generation Perseus model, reducing wait times without compromising output quality.
More consistent drum isolation - improved bleed reduction and cross-section stability, resulting in stems that remain coherent from intro to outro. Andromeda captures more of the original drum content, keeps bass and other instruments out of the stem more effectively, and holds that quality more evenly across the entire track.
Real-World Impact: Music Restoration After a Studio Fire
The practical value of high-fidelity stem separation goes well beyond remixing and production. When a Nashville-based producer lost recordings in a studio fire, LALAL.AI became a critical recovery tool, allowing stems to be isolated from the only surviving mixed files and giving lost performances a second life. Cases like this illustrate why separation quality is not merely a technical metric but a meaningful creative and archival capability.
With the extended Andromeda model, LALAL.AI users gain an A-powered stem separation technology that handles more of these high-stakes scenarios with greater precision, whether they are recovering irreplaceable recordings, extracting clean stems for a remix, or studying the architecture of a finished mix.
Industry Recognition Across Leading Benchmarks
In December 2025, Meta's independent stem separation benchmark ranked LALAL.AI as the top commercial technology in the Pro category.
MusicRadar, in its independent review of stem splitters, awarded LALAL.AI a 5 out of 5 score for both vocal and drum extraction - the two categories most critical to producers, engineers, and musicians working with modern recordings.
At the 2026 Webby Awards, LALAL.AI was recognized in two categories within Apps, Software & Immersive: reaching the Top 5 for Best Use of AI and Machine Learning, and earning Honoree status in Creator, Creative and Media Tools.
About LALAL.AI
LALAL.AI is an AI-powered audio separation platform used by more than 6 million music producers, engineers, DJs, podcasters, and media professionals worldwide. The platform enables fast, high-quality extraction of vocals, instrumental, drums, bass, guitar, synth, and other stems from any audio or video file. LALAL.AI is available via web, desktop, and mobile applications, as well as through API and VST plugin integrations for professional workflows.
Media Contact:
Klara Alexeeva
PR & Comms, klara.alexeeva@lalal.ai
SOURCE: LALAL.AI
View the original press release on ACCESS Newswire:
https://www.accessnewswire.com/newsroom/en/computers-technology-and-internet/following-top-industry-benchmarks-lalal.ai-expands-andromeda-with-1154849
