Any song.
Any voice.
One tap.

VocalMe is an AI-powered cover-song generator. You pick a song — from the in-app library, an MP3 you upload, or a track from your device — and choose a voice from the model library or one you've trained yourself. The AI extracts the vocal, generates a new performance in the chosen voice using the original melody and rhythm, and recombines it with the instrumental. The result is a 'cover' where the singer has been swapped but the song stays the same. AI Splitter, music video creation, and reference-audio AI music generation are layered on top of this core flow.

Honest answer — quality varies a lot. For viral TikTok content where the joke is 'X sings Y,' it works well and the stylized AI texture is part of the appeal. For straight covers you'd want to release seriously, the quality often falls short. App store reviews regularly report robotic or distorted output, especially on custom voices trained from limited samples. Stylistically distinctive voices in the library (high pitch, heavy vibrato, breathy delivery) tend to render better than neutral ones. Pop and hip-hop genres convert more cleanly than country or jazz. Run a 15-second preview before committing to a full song.

Yes. Custom voice training is one of the headline features — record or upload audio samples of yourself singing or speaking, and VocalMe builds a voice model from them. More samples and higher-quality recordings produce better models. Singing samples in your natural range work best. The trained voice becomes available across all your covers. Caveat: many users report that custom voices need significantly more training data than the app initially suggests to sound natural, and the first version of a trained voice often needs retraining with additional samples before it's usable.

AI Splitter isolates vocals from instruments in any song you upload. The output is two stems — vocal-only and instrumental-only — which you can use to build karaoke versions, remix at the stem level, sample the vocal for new tracks, or just listen with the original vocal stripped out. It's the same technology used inside the cover-song generator, exposed as a standalone tool. Quality is competitive with other consumer splitters for most modern pop, hip-hop and electronic recordings. Older recordings, dense rock mixes, and tracks with heavy reverb tend to produce more noticeable artifacts.

Partial. Currently, full track export as MP3 or WAV is available within the app. However, downloading isolated vocal stems directly is more limited than it was in earlier versions — some users have reported needing to use screen recording or external conversion workflows to get the exact audio file format they need, especially when separating just the AI-generated vocal from the mix. If stem-level export is critical to your workflow, check the current export options against your needs before subscribing to a long-term plan.

VocalMe uses a subscription model with two main tiers — weekly at around $9.99/week or yearly at significantly lower effective monthly cost. There's also a custom voice training fee in some plans on top of the base subscription. The paywall appears very early in the onboarding flow, before you've made a cover, which surprises a lot of users. The free tier lets you preview the core workflow but most actual cover generation, custom voices, and AI Splitter access requires a paid plan. Cancel anytime through App Store / Google Play subscription settings.

Complicated. The mechanical and synchronization rights for the underlying song still belong to the songwriter and publisher — TikTok and Instagram have broad music licensing deals that cover most casual social posts, which is why most AI covers stay up. YouTube uses Content ID, which may flag or monetize-claim your video back to the original rights holders. Cloning a famous singer's voice raises separate publicity and likeness questions that vary by jurisdiction. For personal, non-commercial social posts of well-known songs, you're usually fine in practice. For commercial release, paid sync placement, or covers using cloned celebrity voices, get legal advice.

Different tools for different jobs. Suno generates full original songs from a text prompt — VocalMe needs an existing song and just changes the voice. BandLab's 15 AI voices work inside a full DAW with multitrack editing, while VocalMe is a single-purpose cover generator with simpler editing. LANDR focuses on mastering and distribution, not voice generation at all. If you want viral one-tap covers for social, VocalMe is purpose-built. If you want to record original tracks with AI vocal options, BandLab. If you want full AI song generation, Suno. If you want to master and release your own music, LANDR.

VocalMe is developed by Talent Me Technology Pte. Ltd., listed in app stores under the publisher name Luckon. Talent Me is registered in Singapore and operates a small portfolio of consumer AI music apps. The company is relatively new and not as established as legacy music-tech players like LANDR (Montreal, 2014) or BandLab (Singapore, 2015), which is reflected in both the rapid feature iteration and the inconsistent quality reports across versions. Support contact is [email protected] per the developer response in the App Store. The privacy policy is hosted at d1e0dtlz2jooy2.cloudfront.net/inter-web/vocalme/privacy.html.