API Overview & Authentication

OneASR provides a standardized, high-performance REST and WebSocket API suite for speech recognition, media parsing, and file management.

---

Base URLs

• Local Deployment: http://localhost:8000
• OpenAI Compatible Endpoint: http://localhost:8000/v1
• Realtime WebSocket Endpoint: ws://localhost:8000/v1/realtime (or /v1/realtimeext)

---

Authentication

All requests to OneASR endpoints require a valid API key configured in config.yaml (api_key: sk-oneasr-...).

You can pass the API key using any of the following three methods:

1. HTTP Bearer Token (Standard OpenAI Format)

Authorization: Bearer sk-oneasr-v1-p_L3kXm9QZ8sT2vA4wE7rY1u

2. Custom Header (X-API-Key)

X-API-Key: sk-oneasr-v1-p_L3kXm9QZ8sT2vA4wE7rY1u

3. Query Parameter (Recommended for WebSockets)

ws://localhost:8000/v1/realtime?api_key=sk-oneasr-v1-p_L3kXm9QZ8sT2vA4wE7rY1u

---

Functional API Groups

Group	Endpoints	Protocol	Description
OpenAI Audio	/v1/audio/transcriptions, /v1/models	HTTP REST	Drop-in OpenAI speech recognition and model discovery
Realtime Stream	/v1/realtime, /v1/realtimeext	WebSocket	Low-latency duplex live audio streaming with VAD
File Transcription	/v1/file-transcription/transcriptions	HTTP & SSE	Async background processing for large media files (up to 2GB)
File Storage	/v1/files/*, /v1/files/fetch-url	HTTP REST	File uploads, storage management, and remote URL downloads
Media & Engine	/v1/media/parse, /v1/providers	HTTP REST	Media track extraction and engine provider status