Detect AI voices & synthetic audio
Drop an MP3 or WAV and watch the waveform render in your browser. Choose whether to check for AI-cloned speech or AI-generated music, then run the VERITAS Voice Engine — it returns a clear verdict with a forensic confidence score.
VERITAS Engine onlineMP3 · WAVVoice + music detectionZero storage
Drag your audio file here
or tap to browse · MP3 & WAV · max 25 MB
VERITAS Voice Engine
Select whether to check for AI-generated speech or AI-generated music, then drop your file.
Neural TTS & voice cloning
AI-generated music
Vocoder & resynthesis traces
Only the selected 10-second window is sent to the API. Audio is never stored.
Inside the VERITAS Voice Engine
How audio forensics work
How it works
- 1Drop your fileDrop or browse an MP3 or WAV. The waveform decodes in your browser — you can see it render before anything is submitted.
- 2Select your clipIf your file is longer than 10 seconds, drag the highlighted window on the waveform to choose exactly which 10-second segment gets analyzed.
- 3Choose voice or musicPick whether you want to check for AI-generated speech or AI-generated music, then hit Analyze.
- 4Read the verdictYou get an AI probability score, a confidence level, and a plain-language summary of what the engine found.
What we detect
- Neural TTS & AI voice cloning
- Vocoder & resynthesis artifacts
- AI-generated music & instrumentals
- Synthetic envelope patterns
Transparent results
Every result shows the AI probability, a confidence level, the analyzed clip duration, and a plain-language explanation of what the engine found — not just a yes or no.
Confidence scoreContent-type labelZero storage