Upload two audio files and get a deep forensic comparison. The AI analyzes melodic similarity, harmonic content, production techniques, and structural patterns to determine how alike two tracks are and whether either shows signs of AI generation. Useful for detecting plagiarism, verifying remix authenticity, or comparing original and sampled material.
Melodic similarity compares the contour, interval patterns, and rhythmic profile of the most prominent melodic line in each track. Two tracks can have the same key and genre without being melodically similar. The tool looks specifically for shared patterns in the melody itself: the sequence of relative pitch changes and rhythmic values that define a musical phrase, which is the most common form of musical borrowing.
The tool provides an evidence-based similarity score and a detailed breakdown of shared elements, which can inform a plagiarism claim. However, legal determination of music plagiarism requires expert testimony and judicial or arbitration proceedings. The analysis gives you an objective starting point for deciding whether to pursue a claim or consult a music attorney, but is not itself a legal verdict.
Similarity is reported across multiple dimensions. A harmonic similarity score above 70% combined with a melodic similarity above 60% and structural alignment are meaningful indicators of potential copying. Similarity in any single dimension alone is less significant. Many songs share the same chord progressions or similar tempos without raising copyright concerns, since these elements alone are generally not protected.
Yes. If you've obtained a license to create a remix and want to confirm that your new version is sufficiently transformative, this tool can help you understand how much of the original material persists in your remix. A high similarity score may indicate you've borrowed too heavily and should rework sections to create more distance from the source. A low similarity score suggests your remix has enough original transformation to stand independently.