Developer platform

Audio authentication, through one API.

Send audio and get a result with its score, decision threshold and model version. Aletheia checks whether speech is synthetic. Eidos checks whether a voice matches a speaker you enrolled.

Create an accountSign inRead the docs

New accounts receive 100,000 credits once the email address is verified. Each model operation uses 50 credits.

Two models, one key.

AletheiaDeepfake detection

Is this a real voice?

Input
Live or recorded audio
Output
Synthetic score + flagged segments
Endpoint
POST /v1/deepfake/detect
Cost
50 credits per successful operation

EidosSpeaker verification

Is this the right person?

Input
Caller audio + enrolled voiceprint
Output
Speaker match score
Endpoint
POST /v1/speakers/enroll · /v1/speakers/verify
Cost
50 credits per successful operation

A score is evidence, not proof. Use it alongside your own checks, and keep a person in the loop for decisions that matter.

How it works.

  1. Create an API key

    Pick only the scopes it needs. The full key is shown once, then only its prefix.

  2. Send audio

    WAV, MP3, M4A, FLAC, OGG or WebM, up to 20 MB. Send an Idempotency-Key so a retry is never charged twice.

  3. Read the result

    Every response carries the score, the threshold it was compared with, the model version and a request ID.

Example request
curl https://platform.detectif.ai/v1/deepfake/detect \
  -H "Authorization: Bearer $DETECTIF_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -F "file=@call.wav"

Integration and billing

Use it from an agent

The same operations are available over MCP at /mcp, with the same keys, scopes and credits.

MCP setup

Billing you can audit

Credits are reserved before inference and charged once, only when the model returns a result. Rejected or failed requests are released. Every operation is listed with its request ID.

Create an account