Voice Authenticity FAQ
Common questions about ORAVYS voice authenticity analysis, forensic voice technology, deepfake detection, privacy, and pricing.
General Questions
Q What is voice authenticity analysis?
Voice authenticity analysis examines the acoustic properties of speech to determine whether audio is genuine human voice or synthetically generated. It also detects acoustic inconsistencies between different dimensions of the voice signal. ORAVYS uses multi-signal forensic analysis of acoustic features across multiple dimensions per audio frame to provide complete voice authenticity assessments.
Accuracy & Technology
Q How accurate is ORAVYS deepfake detection?
ORAVYS does not publish an accuracy figure. Every paid result falls in one of three bands: strong synthetic signals, no strong synthetic signal, or inconclusive. Results are indicative, not definitive, and a recording that does not support a firm call is reported inconclusive.
Q What audio formats and quality are required?
ORAVYS accepts common audio formats including WAV, MP3, M4A, OGG, and WebM. For optimal results, recordings should be at least 5 seconds long with minimal background noise. The platform includes built-in voice isolation to handle noisy environments. Audio is internally resampled to a standard analysis rate. You can also record directly in the browser using the recording interface.
Q What are the ORAVYS checks?
The checks are independent analyses of the recording. Each check looks at a different signal and contributes to the final assessment, which is one of three results: strong synthetic signals, no strong synthetic signal, or inconclusive.
Privacy & Compliance
Q Is my voice data kept private?
ORAVYS takes privacy seriously. Audio samples are processed in secure cloud infrastructure hosted on Google Cloud Platform and are not shared with third parties. Users can request deletion of their data at any time. All processing complies with GDPR requirements. The platform does not store audio longer than necessary for analysis completion, and voice data is never used to train models without explicit consent.
Q Is ORAVYS compliant with the EU AI Act?
ORAVYS supports Article 50 documentation. This is our own assessment; it has not been reviewed by a regulator or outside counsel. The platform provides probabilistic assessments and never makes deterministic claims about individual behavior. Paid results fall in one of three bands, with a synthetic-voice score shown when the result is conclusive, and clear disclosure of what the result does and does not prove. ORAVYS is a decision-support tool for qualified professionals, not an automated judgment system.
Pricing & Access
Q How much does ORAVYS cost?
ORAVYS offers a free daily check, a single scan at $9.90, a 5-scan pack at $39.90, and an API plan at $99/month for regular volume (100 scans, API access). Business, Enterprise, and academic pricing with custom SLAs is available on request. Visit the pricing page for full details.
Q Is there an API for developers?
Yes. The ORAVYS API is included in the API plan. It supports API-key authentication, tier-based rate limiting, and returns structured JSON with the verdict, label, and a synthetic-voice score when the result is conclusive. A Python SDK is also available. See the contact page to request API access.
Q Can I use ORAVYS for academic research?
Absolutely. ORAVYS offers an Academic API program with discounted rates for university researchers. The platform's standardized multi-engine output provides rich feature sets for studies in linguistics, communication sciences, and speech pathology. Contact us for academic pricing and batch processing capabilities.
Still Have Questions?
Our team is here to help. Reach out for technical questions, enterprise inquiries, or partnership opportunities.
Contact Us Try a Free Analysis