For banks Voice anti-deepfake
Enterprise

Voice authentication with deepfake detection

We detect whether a voice is human or AI-generated and return a confidence score. Simple REST API, credit-based billing, and clear results for fraud and risk teams.

Sign in / Create account Sandbox available
~1s
Typical latency*
API
Simple REST
Credits
Pay-as-you-go

*Depends on audio length and queue.

How it works
  1. Your app calls /analyze with a valid X-API-Key and an audio file.
  2. We validate the key, analyze the audio, and compute “human vs. synthetic”.
  3. You receive the score, signals, and a reference_id for audit.

No valid key → 401. No credits → 402. Nothing is processed without a valid key.

Why banks use VoiceScreen

Lock at the door

API key validation happens before any processing. Without a valid key, nothing runs.

Logs & audit

Every analysis has a reference_id for traceability and audit trail.

Credits with control

Per-user wallet, purchase & usage ledger, and a billing kill switch in the database.

Pricing

Starter

US$ 50

1,000 analyses • US$ 0.05 per analysis

  • API key per customer
  • Credit-based billing
  • HTTPS & authentication
  • No data retention (we do not store audio)

Need more volume? Pro and Enterprise API plans coming next.

Enterprise Annual Plans

Silver

Up to 1,000,000 analyses/year • SLA 95% • Basic support (email)

US$ 75,000 / year
  • Contract + SLA
  • Compliance package available
  • Simple REST API

Buy now — Silver

Gold

Up to 10,000,000 analyses/year • SLA 99% • 24/7 support • Dedicated integration

US$ 250,000 / year
  • Contract + SLA
  • Dedicated integration channel
  • Compliance & audit support

Buy now — Gold

Platinum

Unlimited usage • Premium support • Anti-fraud consulting included

US$ 500,000+ / year
  • Contract + highest SLA
  • Premium success & consulting
  • Custom security reviews

Buy now — Platinum

Need something different (e.g., 2M analyses/month)? We offer a custom Enterprise contract.

Integrate in minutes

curl -X POST "https://api.voicescreen.cloud/analyze?reference_id=demo123" \
  -H "X-API-Key: YOUR_KEY_HERE" \
  -F "file=@client.wav"

Response: 200 JSON on success; 401 without key; 402 without credits.

Security & Compliance

Audio data

Transient processing. You decide your analysis log retention.

Encryption

TLS 1.2+ in transit. Secrets in Google Secret Manager.

Regulatory

SLA and DPA available for banks and insurers.

Accuracy & Continuous Improvement

Our current detection accuracy is 95%, based on more than 300,000 voice samples tested. These tests include human voices across multiple languages and emotional states, as well as synthetic voices created by different AI laboratories.

However, just like new types of computer viruses keep appearing, synthetic voices also evolve. We continuously retrain our model to keep pace with new generations of voice synthesis technologies.

If you find a synthetic audio that was not recognized as such, please help us improve by submitting it:

Submit a synthetic sample (not recognized)

FAQ

Which audio formats? WAV/MP3/OGG up to 20 MB.

Typical response time? Around 1 second per typical analysis.

How do I buy credits? Stripe Checkout (test/production). Credits are added instantly after payment.