Song key finder. Upload audio and AI detects the musical key, chord progression, and modulations.
Select the AI model for audio analysis. Different models may have different capabilities.
Record audio directly from your microphone
To find what key a song is in, upload a clip and the AI returns the primary key (major or minor) with a confidence level, the chord progression written out both as chord names and Roman numerals, any key changes or modulations with where they happen in the track, and the Camelot wheel notation DJs use for harmonic mixing. It also flags borrowed chords and unusual harmonic choices, which is where most automatic key detectors fall flat: a song that leans on a flat-VI or mixes modes will confuse a pure pitch-class algorithm, but the analysis here explains what the harmony is actually doing instead of just printing a letter. That makes it useful in two very different worlds. DJs get the Camelot code (like 8A or 5B) to build harmonically compatible transitions, and musicians get a real harmonic map: what the progression is, how it functions, and where it steps outside the key.
Upload the audio and the AI examines the harmonic content. Key detection identifies the tonal center and quality (major or minor) and attaches a confidence level, committing to its best read on ambiguous material rather than refusing. Chord progression analysis writes out the changes two ways: chord names for players who want to follow along, and Roman numerals for anyone studying how the progression functions. Modulation detection scans the full clip for key changes, noting where they happen (the classic final-chorus lift, a bridge that pivots, or a song that genuinely lives in two keys). Camelot wheel conversion translates the key into DJ notation for harmonic mixing. Finally, the harmonic commentary points out borrowed chords, modal mixture, and any choices that make the song interesting, which is the context a plain key label never gives you.
Upload a clip and you get the key with its confidence level, the chord progression written as both chord names and Roman numerals, any modulations with their locations, and the Camelot code for DJ mixing. The progression is the part most key finders skip, and it is usually what you actually wanted to know.
It is the DJ shorthand for harmonic mixing: every key maps to a code like 8A or 5B, and tracks with adjacent codes blend without clashing. Getting the Camelot notation with the key means you can drop the track straight into a harmonically mixed set without doing the translation yourself.
Yes, modulation detection runs across the whole clip. The truck-driver key change before the final chorus, a bridge that lifts a minor third, a song that genuinely lives in two keys, each gets reported with where it happens. If you only need one section's key, upload just that section for the cleanest read.
Usually the relative major and minor confusion: A minor and C major share every note, so pitch-counting algorithms flip a coin while the actual answer depends on where the music resolves. Borrowed chords and modal songs make it worse. This analysis reasons about the resolution and explains its choice, which is exactly the tiebreaker you need.
Strong on clearly tonal music, candid on the hard stuff. Confidence comes back with every result, and the genuinely ambiguous cases (modal vamps, drone-based tracks, two-chord songs that never resolve) get a best call plus the alternative reading. Heavily detuned or noisy recordings lower confidence, so a clean studio clip beats a concert phone video.
That is the everyday use. The progression arrives in chord-name form you can play immediately, with Roman numerals showing the function so the same song transposes to any key. Borrowed chords and oddities get flagged, which saves you from assuming you misheard when the song really does pull a chord from outside the key.
Voice personality profiler. Upload a clip and AI infers personality traits from how you speak.
Empathy tone detector. Upload a clip and AI rates how warm, attentive, and compassionate your voice reads.
Voice match checker. Upload two clips and AI estimates whether they're the same speaker.
Background noise identifier. Upload audio and AI tells you what ambient sounds are in your recording.
Audio location finder. Upload a clip and AI guesses the room, environment, or country based on acoustic cues.
Spanish pronunciation checker. Record yourself speaking Spanish and AI rates your accent, vowels, and rhythm.