Realtime API

WSS /v1/realtime?model={model}&intent={intent}

WebSocket endpoint for real-time bidirectional audio streaming. Proxies to OpenAI's Realtime API.

Connection

const ws = new WebSocket(
  "wss://api.osmapi.com/v1/realtime?model=gpt-realtime",
  {
    headers: {
      "Authorization": "Bearer $OSM_API_KEY"
    }
  }
);

Note: The gateway adds the OpenAI-Beta: realtime=v1 header automatically.

Query Parameters

Parameter	Type	Required	Description
`model`	string	Yes	`gpt-realtime` or `gpt-realtime-mini`
`intent`	string	No	`transcription` for transcription-only mode

Models

Model	Text Price	Audio Price
`gpt-realtime`	$4/$16 per 1M tokens	$32/$64 per 1M audio tokens
`gpt-realtime-mini`	$1/$4 per 1M tokens	$8/$16 per 1M audio tokens

Transcription Mode

For streaming transcription only (no conversation):

const ws = new WebSocket(
  "wss://api.osmapi.com/v1/realtime?model=gpt-realtime&intent=transcription"
);

Authentication

Pass your OSM API key as a Bearer token in the WebSocket connection headers.

See the Realtime feature guide for detailed usage examples.

How is this guide?