Tool Directory
Browse tools
Tap the handle to open the tool directory.
Tool Directory
Browse tools
Tap the handle to open the tool directory.
Quick access to lyrics-to-melody, vocal synthesis, stem separation & studio music tools.
Tool Directory
Browse tools
Tap the handle to open the tool directory.
Upload a reference, add your lyrics, and hear a new song inspired by your sound.
You already know the kind of music you want. Bring a reference that captures the sound, then add the words you want sung.
Choose a recording that captures the musical character you're looking for. Use your own recording or audio you have permission to process.
What do you want the new song to say? Paste the words you'd like to hear sung.
0 / 5000 characters
Generation Cost: 80 Credits
Ready to Hear Your New Song?
We'll use your selected reference as musical inspiration and your lyrics as the starting point for a newly generated song.
Selected Clip: 30 seconds
Lyrics Language: English
Output: Full Song MP3
This creates a new song inspired by your reference. The result may have a different melody, rhythm, arrangement, and vocal performance.
Select the language of your lyrics. The AI determines the singing from the lyrics themselves, so pronunciation accuracy may vary.
Section labels such as [Verse] and [Chorus] help describe the structure of your lyrics. Exact wording and phrasing may vary in the generated performance.
Have a sound in mind but can't quite describe it?
Upload a short music reference, add your lyrics, and hear AI create a new song inspired by the style and feel of your recording.
Whether it's a demo you've recorded, an instrumental you've produced, or another track you're authorized to use, your reference can help guide the musical direction without starting from a blank page.
Reference audio guides the musical feel. It does not reproduce the source recording or guarantee the same melody, singer, or arrangement.
Sometimes a recording communicates more than a long description ever could.
Listen to how a short music reference can inspire a new song with different lyrics and its own musical interpretation.
Acoustic / Indie
A gentle acoustic demo provides the inspiration for a new song with an intimate musical feel, AI singing, and a fuller arrangement.
The generated song explores the reference's atmosphere without reproducing its original melody.
Piano / Ballad
A reflective piano recording becomes a musical guide for an AI-generated ballad built around newly supplied lyrics.
Similar emotional direction, but a newly interpreted song structure and performance.
Electronic / Pop
Use an energetic electronic reference as inspiration for a new song with contemporary production and AI vocals.
The reference influences musical character rather than acting as a source track for remixing.
Jazz / Soul
A jazz-inspired instrumental reference guides the creative direction of a new song with expressive vocals and a distinct arrangement.
Hear how one musical atmosphere can lead to a different finished song.
Audio examples are being prepared.
Example results demonstrate reference-guided music generation. Actual outputs vary. Only recordings with appropriate usage rights should be submitted.
Try My Own Music ReferenceYou know the music you want to make.
Maybe you've recorded a rough demo on your phone.
Maybe you've produced a short instrumental that captures the mood of your next song.
Or maybe you have an authorized reference that expresses a sound you're struggling to put into words.
You could spend hours trying to explain every instrument, rhythm, and production detail.
Or you could start with the sound itself.
Reference Audio to Song lets you use a short recording to guide a new AI-generated song around the lyrics you've written.
Each generation creates a new interpretation. Exact lyric accuracy, musical similarity, and arrangement details aren't guaranteed.
A reference should help communicate your musical direction, not serve as a request to duplicate another song.
Upload My Music ReferenceNot every music reference serves the same purpose.
Start with an audio recording whose overall musical style you want to explore. The AI interprets the sound and creates a new song around your lyrics.
Best For: Musical mood, general style, production character, and creative direction.
Start with a MIDI melody you've composed. The AI uses that melody as a musical guide for a full song with your lyrics.
Best For: Creators who already have a specific melodic idea.
Start with the words you've written and let AI create the accompanying musical direction.
Best For: Songwriters who have lyrics but no existing reference recording or MIDI melody.
Start with lyrics and create a song, a separated backing track, and a guide vocal.
Best For: Songwriters who want musical accompaniment and vocal guides to develop their material.
If you want a new song inspired by the overall sound of a recording, Reference Audio to Song is the right place to start.
Reference-guided AI music generation is not audio copying.
Your recording provides musical context, but the generated song may use a different melody, arrangement, rhythmic feel, vocal performance, and duration.
The AI may follow some parts of the reference more closely than others.
This workflow isn't designed to create exact covers, preserve every note, copy a singer's voice, or reproduce a copyrighted master recording.
Listen to the generated result and treat it as a new creative interpretation.
A Reference Audio to Song Generator uses a short music recording as creative guidance for a new AI-generated song. You upload an audio reference, add your lyrics, and the system creates a song with singing and instrumental accompaniment.
Yes. You can use your own music recording as a reference, provided it meets the supported format and duration requirements. The result is a newly generated song rather than a direct transformation of your recording.
No. The reference is intended to guide the overall musical direction, not to reproduce the original recording. Similarity may vary, and generating a new song does not guarantee that it will be free of similarities to existing music.
You can upload an MP3 or M4A recording containing music you own or have permission to use for this process. A clear instrumental or musical demo can be a useful starting point.
The generator uses a 30-second reference clip. If your original recording is longer, select the section that best represents the sound you want before generating your new song.
You can prepare a 30-second section from a recording you have the right to process. The generator uses that selected section for inspiration, not the entire song as material to copy.
Only if you have the necessary rights or permission for that use. Being able to listen to, purchase, or download a song does not automatically give you permission to submit it to an AI music service.
Yes. Enter the lyrics you want sung. You can organize your writing with section labels such as [Verse] and [Chorus]. The generated vocal performance may vary in wording, timing, and phrasing.
Yes. This workflow is designed to generate a complete song with an AI singing voice and instrumental accompaniment.
Not in this workflow. The generated song uses an AI voice. Uploading a musical reference does not clone the singer's voice or reproduce the original vocal performance.
Not necessarily. Reference audio primarily guides the musical style and feel. If your goal is to guide the AI with a particular sequence of notes, a MIDI-based Melody to Song workflow is more suitable.
No exact preservation is guaranteed. The model may interpret the reference's rhythm, instrumentation, and production character differently in the new song.
This version requires lyrics. If you want to generate a song from a general written description instead, use a prompt-based music generation workflow.
The workflow is planned around English, Spanish, Japanese, Korean, Portuguese, German, Italian, French, and Russian. Singing accuracy varies by language and by generation. Other languages are not included in this workflow.
This reference-guided version uses your audio recording as its musical direction. It doesn't provide independent genre, mood, or voice-cloning controls. Choose a reference that communicates the sound you want.
Not necessarily. The 30-second limit applies to the reference clip, not the finished song. The model determines the generated song's duration.
Yes. Once your generation completes successfully, you can listen to the finished audio in the result player before downloading the MP3.
Yes. You can generate another interpretation using the same reference and lyrics. Each new generation is a separate paid request and may sound different.
Not from this generator. It delivers a combined song recording. If you need separate vocals and instrumental audio, you can use the site's Vocal Remover workflow afterward.
Commercial use depends on the applicable model, platform, and output terms, along with your rights to the reference audio and lyrics. Generation does not automatically guarantee copyright ownership, exclusivity, or unrestricted commercial rights.
Uniqueness isn't guaranteed. Other users may receive similar musical results, and the output may share stylistic characteristics with existing recordings.
The number of credits required is shown before you start generating. A new version is another paid generation. No credits should be charged merely for selecting and previewing a reference file.
If a generation fails for a qualifying technical reason, the workflow should return the applicable generation credits. The result screen displays whether the refund is completed or still being processed.
No. A remix changes or reworks an existing recording, and a cover reinterprets an existing composition. This tool uses a reference as musical inspiration for a newly generated song with your supplied lyrics.
You've already found the musical direction you want.
Bring your reference audio and lyrics together, and explore what a new AI-generated song could sound like.
Your Reference · Your Lyrics · A New AI-Generated Song
Use recordings you're authorized to process. Generated results may differ from the reference in melody, arrangement, vocals, and style.