Docs
Getting started

Introduction

Two Turkish voice models, one API key. Alania turns text into speech; Yanki turns speech into text.

Alania

Streaming text to speech. POST /v1/audio/speech returns WAV as it is generated; first audio in ~{ttfa} ms.

Yanki

File transcription and live streaming. Drop-in compatible with OpenAI Whisper clients.

Base URLs

ServiceURL
Yanki · RESThttps://voice.patientdesk.ai/v1/audio/transcriptions
Yanki · WebSocketwss://voice.patientdesk.ai/v1/audio/stream
Alania · RESThttps://voice.patientdesk.ai/v1/audio/speech
Alania · WebSocketwss://voice.patientdesk.ai/v1/audio/speech/stream

All requests go over HTTPS and JSON errors follow the OpenAI shape. Each account's quota and concurrency limits are shown in the dashboard.

Next steps

Quickstart

Create a key and send two requests: one speaks, one listens.

Authentication

Bearer keys, and short-lived session tokens for the browser.

Errors and limits

Status codes, error codes, quotas and rate limits.