AI Mashup Maker

Choose two audio references and describe the sound you want. Create a new interpretation of both tracks, rather than a direct overlay of the originals.

Audio that matches a recording in the generation service’s catalog will be rejected, even if you have permission to use it. Use your own original recordings or other audio you have the right to upload.

1Audio A

MP3, WAV, M4A, AAC, OGG or FLAC · up to 50 MB · 5 seconds–8 minutes

2Audio B

MP3, WAV, M4A, AAC, OGG or FLAC · up to 50 MB · 5 seconds–8 minutes

3Music style

Song titleTarget duration10–360 seconds

seconds

Advanced options

Lyrics (optional)Audio influence50%How strongly both audio references guide the new song; this does not set an A/B mixing ratio.

The result is a new interpretation of both references. Original vocals and melodies may change.

Loading…

Choose two tracks and describe the style. Your new songs will appear here.

How to use the AI Mashup Maker

What this song mashup maker actually creates

A traditional mashup often puts a familiar vocal over another song’s instrumental. This tool takes a different approach: two audio references guide the creation of a new song. Describe what you want to borrow from each reference, such as a rhythmic feel, a warmer arrangement, or a contrasting mood. The result is a new interpretation, so the original vocal and melody may change. That distinction matters when you choose your inputs and decide whether the result fits your project.

Use this workflow to explore an arrangement: pair a gentle acoustic reference with an energetic electronic track and ask for a restrained verse that opens into a dance-oriented chorus. Choose which feature should lead each section before writing the brief. For a reflective song, the acoustic reference might guide the opening while the electronic pulse becomes more prominent in the chorus. Listen to both sources first and write down that contrast, so you have a clear musical idea to evaluate after generation.

Choose two references that have a purpose

Start with two different audio selections. You can upload a local file or choose an eligible recording from your cloud audio library. Give each reference a role before submitting: one might supply the rhythmic character while the other suggests instrumentation or atmosphere. Choosing two tracks simply because you like them can leave the direction ambiguous. Listen for a specific relationship you can explain in a sentence, then use that relationship to guide your style description.

The accepted input extensions are MP3, WAV, M4A, AAC, OGG, and FLAC. Each file must be no larger than 50 MB and its decoded audio must be between 5 seconds and 8 minutes long. Play each exported file once to check that it opens correctly and contains the passage you intended to use. Choose a clear passage that represents the idea you want to explore. A long silence, a damaged export, or a recording dominated by unrelated speech is unlikely to be a useful musical reference.

Describe the combination instead of listing adjectives

The style field is required. Write a short musical brief that explains how the references should work together. A practical prompt might read: ‘Use the first track’s steady electronic pulse and the second track’s intimate acoustic atmosphere. Keep the verse sparse and let the chorus feel wider, with warm bass and a relaxed vocal delivery.’ This gives the request a hierarchy: rhythm, atmosphere, arrangement, and performance. Use descriptions such as a soft guitar attack, a steady bass pulse, or a spacious chorus to make the direction concrete.

A useful revision changes one meaningful decision. If a result feels too dense, ask for fewer competing instruments; if the contrast disappears, state more clearly where the change in energy belongs. Choose the priority you want to keep, such as the rhythmic character, and explain which part of the arrangement should develop around it. The style field allows up to 1,000 characters. Spend that space on priorities and exclusions that affect the sound, rather than repeating synonyms for ‘amazing’ or ‘professional.’

Set the target length and audio influence

Choose a target duration between 10 and 360 seconds to guide the scope of the new song. A short target can focus on the central hook; a longer target can give a verse and chorus room to develop. For a video, describe whether the opening should be immediate or gradual, then check the actual file length and ending after generation. Use that completed recording to plan the placement and any final trim in your video editor.

Audio influence controls how strongly both references guide the new song. It is an overall reference influence, rather than an A/B volume balance. To emphasize a particular reference’s rhythm or texture, describe that intention in the style field. When comparing experiments, keep the inputs and brief stable while changing the influence setting, so you can judge whether the resulting interpretation is moving in the direction you want.

Add lyrics when words are part of the idea

You can enter optional lyrics of up to 5,000 characters and a title of up to 100 characters. Supply your own words when the lyrical theme matters more than the source recordings’ words. A concise chorus and a clear verse idea are often easier to assess than a long lyric with several unrelated narratives. Keep the intended mood consistent with your style request, and listen for clear pronunciation, comfortable phrase lengths, and room to breathe between lines in the generated recording.

Leaving lyrics blank does not guarantee an instrumental result. If words are part of your idea, give the verse a concrete scene and let the chorus return to one central thought. For a song about a journey home, the verse could describe the last train while the chorus expresses the anticipation of arriving. Read the words aloud before submitting, shorten crowded lines, and use the completed vocal to decide which phrases work well with the musical pacing.

Listen, compare, and keep the useful result

After submission, follow the task in the tool and listen to the available completed results. Compare the arrangement, the relationship between the references, vocal intelligibility, and any abrupt changes. A result that has the right overall mood may still need a different opening or a simpler chorus for your use case. Download the version you want to keep as an MP3 and assess it on the device or in the context where you expect to use it. The page also provides task history for revisiting available recordings.

Reusing settings is useful when you want to refine a brief without retyping everything. Check that the references are still available before you submit another request; an unavailable source must be selected again. Each new submission creates a separate generation task, so decide on the change you want to hear before sending it. Keep a local note of the prompt and settings that produced a promising version, particularly when you are comparing several interpretations of the same pair of tracks.

Check access and recording rights before generating

Sign in to upload your references and create a song. Every account can generate when it has at least 20 usable credits, including credits received at registration or through daily check-in. Each request costs 20 credits, regardless of the chosen target duration. Registration and check-in credits can be used directly, without a subscription or previous purchase. Check your available balance before submitting, especially if you have used another tool since opening this page. If the balance cannot be confirmed, wait and try again. Once a request is running, follow its status rather than submitting the same idea repeatedly. The task history helps you compare the result with your next idea before spending credits on another version.

Use recordings you own or have permission to upload and process. Generating a new arrangement does not automatically grant rights to publish or commercially use the reference music, lyrics, or recording. Before publishing, check the permissions relevant to your intended use. If you want to prepare a shorter source passage, separate a vocal for another project, or create accompaniment around a vocal recording, the related tools below offer different workflows; choose the one that matches the task rather than expecting every operation from a mashup generator.

For a separate project, isolate vocals or accompaniment with Vocal Remover.

Prepare a shorter reference passage with Audio Cutter.

To build accompaniment around a vocal rather than blend two references, try Add Instrumental.

Questions about making a song mashup

How do I start a mashup?

Sign in, choose two different audio references, enter a style description, and select a target duration. Check that at least 20 usable credits are available before submitting.

Will the original singer and melody stay unchanged?

Not necessarily. The tool generates a new interpretation; the original vocal and melody may change. It is not a direct overlay of the recordings.

Does audio influence set the balance between the songs?

No. It controls the influence of both references on generation, not their individual playback levels or an A/B mixing ratio.

Can I make a mashup without writing lyrics?

Yes. Lyrics are optional, but the style description is required. Leaving lyrics empty does not guarantee an instrumental result.

How do I download a completed song?

Open a successful task, listen to the available results, and use the download control for the version you want to save as an MP3.

Can I use check-in credits without a paid account?

Yes. Registration and daily check-in credits count toward your available balance. Any signed-in account with at least 20 usable credits can generate; each request costs 20 credits.

AI Mashup Maker

Choose two audio references and describe the sound you want. Create a new interpretation of both tracks, rather than a direct overlay of the originals.

1Audio A

MP3, WAV, M4A, AAC, OGG or FLAC · up to 50 MB · 5 seconds–8 minutes

2Audio B

MP3, WAV, M4A, AAC, OGG or FLAC · up to 50 MB · 5 seconds–8 minutes

3Music style

Song titleTarget duration10–360 seconds

seconds

Advanced options

Lyrics (optional)Audio influence50%How strongly both audio references guide the new song; this does not set an A/B mixing ratio.

The result is a new interpretation of both references. Original vocals and melodies may change.

Loading…

Choose two tracks and describe the style. Your new songs will appear here.

How to use the AI Mashup Maker

What this song mashup maker actually creates

Choose two references that have a purpose

Describe the combination instead of listing adjectives

Set the target length and audio influence

Add lyrics when words are part of the idea

Listen, compare, and keep the useful result

Check access and recording rights before generating

For a separate project, isolate vocals or accompaniment with Vocal Remover.

Prepare a shorter reference passage with Audio Cutter.

To build accompaniment around a vocal rather than blend two references, try Add Instrumental.

Questions about making a song mashup

How do I start a mashup?

Sign in, choose two different audio references, enter a style description, and select a target duration. Check that at least 20 usable credits are available before submitting.

Will the original singer and melody stay unchanged?

Not necessarily. The tool generates a new interpretation; the original vocal and melody may change. It is not a direct overlay of the recordings.

Does audio influence set the balance between the songs?

No. It controls the influence of both references on generation, not their individual playback levels or an A/B mixing ratio.

Can I make a mashup without writing lyrics?

Yes. Lyrics are optional, but the style description is required. Leaving lyrics empty does not guarantee an instrumental result.

How do I download a completed song?

Open a successful task, listen to the available results, and use the download control for the version you want to save as an MP3.

Can I use check-in credits without a paid account?

Yes. Registration and daily check-in credits count toward your available balance. Any signed-in account with at least 20 usable credits can generate; each request costs 20 credits.