Turn your story into a song that sounds like you
A lyric in your notes, a story you want to tell, or a chorus you have not written yet can be the beginning. Bring a song idea and a trained singing voice to the same music workspace, where you can develop the words, choose a musical direction and hear an original vocal song.
The process combines song generation with voice conversion: the original song supplies the melody, timing and vocal performance, while your saved voice shapes the vocal timbre. You do not need to record a complete performance of every new song. Train your voice once, then reuse it as you explore new ideas.
From your idea to a song in your voice
Choose a trained voice
Select an available official voice or a ready voice from your private library. To use your own voice, upload or record clean solo singing, name it and wait for training to finish. Ready voices can be reused for later songs.
Describe the music
Describe the genre, mood, instruments and story you have in mind. Paste finished lyrics or start with a short idea. Your brief, lyrics and selected voice travel together into the music workspace.
Create and shape the song
Develop an original song and apply your selected singing voice to its vocals in the music workspace. Keep the generated source and your vocal version together, listen to both and refine the next take.
Shape the music, the words and the voice
Give your song a clear direction
Make your words the center of the song
Hear the song through your saved voice
A voice for the songs you want to write
Use your saved singer to explore a new idea, tell a personal story or try a musical direction you have never recorded.
Songwriting demos
Hear your lyrics in an original arrangement with a vocal sound based on your own recordings. Compare a softer verse, a bigger hook or another genre before deciding which version to develop further.
Personal songs and gifts
Turn a shared memory, a thank-you or an anniversary story into a song. Include the details only you know, choose a fitting mood and use your trained voice to make the vocal version more personal.
New styles with a familiar voice
Explore how an acoustic, pop or soul idea works with the same saved singer. Change the musical brief between songs while reusing your voice, and listen for the melodies and arrangements that suit it.
Start with a brief you can make your own
Describe the sound and the story together. These examples are starting points for your song description; add finished lyrics in the separate lyrics field if you have them.
Make a warm acoustic pop song about finding my way home after years away. Use fingerpicked guitar, soft drums and an intimate verse that opens into a hopeful chorus. Keep the vocal melody in a comfortable mid-range.
A clear mood, instrument palette and vocal direction give the workspace a useful starting point.
Write a gentle piano-led song for our anniversary, about the little everyday moments that make a life together. Keep the verses conversational and build to a heartfelt chorus, with light strings near the end.
Add names, memories or your own lines to make the story specific to the person you have in mind.
Turn my lyrics into a laid-back soul song with warm bass, electric piano and a restrained drum groove. Let the verses feel close and relaxed, then give the chorus a stronger melodic lift without rushing the words.
Paste your lyric into the lyrics field, then use the brief to describe its musical setting.
Give your song and voice a strong starting point
The recording teaches the voice's sound, and the generated vocal supplies the performance. A clear brief and clean training material help both parts of the workflow.
Record clean solo singing
Upload or record 2–30 minutes of your voice, without backing music, harmonies or other singers. Use a quiet space and a consistent microphone distance so the training material stays focused on your vocal sound.
Include your natural range
Aim for 10–20 minutes covering comfortable low, middle and higher notes, plus softer and stronger phrases. Choose material that reflects how you sing rather than forcing notes outside your usual range.
Describe the performance as well as the genre
Alongside style and instruments, mention whether you want an intimate, energetic or relaxed vocal. The generated source shapes the melody, rhythm and expression that the converted version follows.
Give your lyrics a readable structure
Separate verses and choruses with line breaks, and mark sections such as [Verse] and [Chorus] when useful. Keep the song description in its own field so the lyrics contain the words you want sung.
Listen to the source performance
Check the generated melody, phrasing and pronunciation before judging the vocal version. If a line feels rushed or a melody does not suit your voice, refine the song direction for your next take.
Compare, then refine
Listen to the generated source and your voice version side by side. Use what you hear to guide the next brief, lyric revision or vocal conversion, while keeping the earlier tracks in your workspace.
Creating AI music with your voice
Choose an available official voice or train a singing voice from your own recordings. Select it here and describe an original song or add lyrics. Your brief and voice are attached to a music workspace, where song generation creates the source performance and voice conversion makes the version with your selected voice.
Yes. Select an available official voice or a private voice with completed training. You can upload or record vocals and train your own voice in the library above. Training is a separate step and does not finish instantly when you upload a recording.
No. Your training recordings establish the saved voice, while the generated song supplies the melody and vocal performance for conversion. You do not need to record a complete take of every new song or sing each lyric revision yourself.
Use 2–30 minutes of clean solo singing without backing music, other voices or heavy effects. For singing, 10–20 minutes across your comfortable vocal range is the recommended starting point. You can record directly in the library or upload an audio file.
Yes. Paste your lyrics into the optional lyrics field and use the song description to explain their musical setting. Verse and chorus labels can help communicate the structure. You can also leave the lyrics blank and develop the words from your idea in the music workspace.
Yes. Available official voices and ready private voices can be selected for later songs and voice conversions without training again. Your private voices stay in your account library. Song generation and conversion still use credits separately.
The training recording is used for the voice's sound. This page starts a new song from your brief or lyrics, so its melody comes from the generated song rather than copying the training recording. If you want to build a song around a melody you hum, use Hum to Song.
This page starts an original song from a brief or lyrics. If you already have a recorded song and want a new vocal timbre while keeping its melody and lyrics, use AI Cover Generator.
Results depend on the training recordings and the source vocal performance. Listen to the result, consider the voice's natural range, and adjust the song or conversion pitch when needed.
Voice training, music planning, song generation and voice conversion are charged as separate steps. The button above shows the starting music-planning message cost; it is not the total price of a completed song in your voice. Creating a voice does not include unlimited songs or conversions.
Build your next vocal workflow
What would your next song sound like?
Choose your trained voice, bring a story or a lyric, and start shaping an original song in your music workspace.