A little help goes a long way
From your first beat to voice swaps and downloads. Here are the answers to the questions that come up along the way.
Choose an artist style or describe the sound you have in mind, then generate. You can make a full song with vocals or choose an instrumental beat. Start with a genre, mood, and a few details about the sound; you do not need recording equipment or music theory to try it.
Yes. The artist catalog includes styles such as Drake, The Weeknd, Taylor Swift, and Justin Bieber. An artist style guides the production and musical feel of a new AI song. It does not make the song an official recording by that artist, and selecting a style does not automatically clone their voice. ReplicaBeats is not affiliated with or endorsed by the artists listed.
Yes. The song generator has vocal and instrumental modes, along with custom lyrics, titles, and generation controls on every credit plan. You do not need Pro to write your own song. Use lyrics you wrote or have permission to use; generating the audio still uses credits.
New accounts receive 5 welcome credits. A standard song generation uses 1 credit, and a voice conversion uses 5 credits. Generating a song and then converting its voice are separate actions, so that combination normally uses 6 credits. Some browser audio tools are free to use without spending credits. Paid credit plans add to your balance each month and include commercial use under the paid-plan terms.
Voice Swap converts the singing performance on an existing track into a selected voice model, then combines the converted vocal with the backing track. You can upload audio or use a supported YouTube link in the Voice Swap console, or choose a song from your library in the singing-voice tool. Use source audio and voices you have the necessary permission to use.
No. Artist styles and singing voice models are separate catalogs. A style can be available for generation even when that artist has no available voice model. Voice conversion also depends on the voice engine being online. Check the voice selector for the models currently offered.
The free vocal remover runs in your browser and estimates vocals and instrumentals from a stereo track. It is useful for a quick karaoke or practice version, but centered instruments can be affected too. The stem splitter offers server-processed separation options for cleaner vocals or separate instruments. Available stems, processing time, and credit costs depend on the option you choose.
Song generation usually takes a few minutes. Voice conversion can take longer because the vocals need to be separated, converted, and mixed. Timing depends on the track and the processing queue. Follow the status shown in the tool; an early audio preview may appear before the final downloadable track is ready.
Finished songs are saved to My Songs, where you can listen and download them. Star the tracks you want to keep in your library: unstarred songs are subject to the 30-day retention policy. Download a local copy of important tracks as well. If you closed the tab during generation, return to My Songs so the app can check for completed or failed jobs.
The Terms of Service give paid plans a royalty-free commercial license for generated audio; free-plan songs are for personal, non-commercial use. That license does not clear someone else's lyrics, recordings, likeness, or voice rights. Before releasing or monetizing a track, make sure you have the needed permissions and meet the rules of the platform where you publish. A subscription does not guarantee that every voice-swapped cover is cleared for commercial release.
The app is designed to return the credits charged for a failed generation or voice conversion after the failure is confirmed. A job that is still processing is not treated as failed just because it takes longer than expected. Check My Songs and your credit activity before starting the same job again. If a confirmed failure has not been credited back, contact support with the job details.