AI Voice Detector

AI Voice Detector provides a verdict and an AI probability score for each check, using a single model behind both the web detector and API. It supports uploading voice notes, call recordings, or clips and returns a four-step verdict with segment scores.

The tool is designed for use by newsrooms, banks, legal teams, investigators, and developers who need a clear answer about whether a recording is machine-generated. It emphasizes post-hoc analysis of recordings rather than live monitoring.

Users can save results to history and share links that do not expire, and the API allows programmatic access to verdicts, scores, and per-segment results.

The detector works on synthetic and stock voices alike and provides a description of what the verdict means, including a four-step scale and an AI probability score.

Use cases

• Newsrooms check suspicious recordings before publication and keep the result alongside reporting.
• Banks and call centers screen flagged recordings and file the result in a case file or for a law-enforcement referral.
• Developers embed the detector into their own tooling via the API.