API Documentation

Integrate VCaaS voice cloning and deepfake detection directly into your own applications using our REST API.

Authentication

All API endpoints require a Firebase ID token passed in the Authorization header.

Authorization: Bearer YOUR_FIREBASE_ID_TOKEN

Generate Voice Clone (Zero-Shot)

Generates synthetic speech using a reference audio file. The output is automatically cryptographically watermarked.

POST/api/v1/tts/clone

Request (multipart/form-data)

text (string, required): The text to synthesize.
reference (file, required): A clean 5-10 second .wav or .mp3 sample of the target voice.
language (string, optional): Language code (e.g., 'en'). Defaults to 'en'.

curl -X POST \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "text=Hello, this is a cloned voice." \
  -F "language=en" \
  -F "reference=@sample.wav" \
  https://vcaas.onrender.com/api/v1/tts/clone --output result.wav

Deepfake Verification (6-Layer)

Runs an audio file through our full 6-layer defense system (anti-spoofing, prosody, semantic coherence, and cryptographic watermark detection).

POST/api/v1/verify/full-analysis

Request (multipart/form-data)

file (file, required): The audio file to analyze (.wav, .mp3).

curl -X POST \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "file=@suspect_audio.wav" \
  https://vcaas.onrender.com/api/v1/verify/full-analysis

Response

{
  "is_authentic": false,
  "confidence": 1.0,
  "reason": "Watermark detected: Synthetic audio",
  "layers": {
    "layer1_antispoof": { "passed": false, "score": 0.12 },
    "layer6_watermark": { "passed": false, "watermark_present": true }
  }
}