Hear a real AI song mashup
Compare the two source recordings with both outputs from one generation. These are complete audio files from our test, with no added soundtrack or edited highlight reel.
Tested on by AI Song Mashup. The source uploaders list both recordings under CC0. The generated tracks are new interpretations, not exact vocal-over-instrumental edits.
Reference A: An Untold Legend
SkyleTheFrench · original recording · 2:01
Reference B: Cozy Puzzle In-Game 2
MintoDog · original recording · 2:08
Generated take 1
AI-generated interpretation · 3:13
Generated take 2
AI-generated interpretation · 3:20
What happened in this test?
| Setting or result | Observed value |
|---|---|
| Input | Two MP3 references, shown above |
| Generation | Kie mashup API, requested model V6 |
| Requested duration | 30 seconds |
| Actual take 1 | 193.224 seconds |
| Actual take 2 | 199.632 seconds |
| Output audio | MP3, stereo, 48 kHz |
Both files downloaded and decoded completely. The requested duration was not followed in this run. A target length is a request, not a guarantee; trim the result in an editor if you need an exact running time.
How should I judge the result?
Listen to each reference before the outputs. Compare the arrangement, rhythmic character and melodic ideas. Decide whether the new piece is useful for your project rather than assuming that every part of either recording will be preserved.
This is one successful provider test. It does not establish a typical success rate, consistent musical quality, exact voice preservation or a completed customer payment flow. Public generation and checkout are being prepared; the studio shows their current availability.
Which workflow fits your music?
For a new interpretation, follow our guide to mashing up two songs with AI. If you need an unchanged vocal or exact beat alignment, read the AI fusion versus stem mixing comparison.