Skip to content

VoiceOpenAI

OpenAI launches the Realtime API for speech-to-speech apps in public beta

Realtime API at a glance

The Realtime API let paid developers build low-latency voice conversations on GPT-4o over a single WebSocket connection, with function calling, at about $0.06 per minute of audio input and $0.24 per minute of output.

What changed with Realtime API

At DevDay on October 1, 2024, OpenAI launched the Realtime API in public beta for all paid developers. Powered by gpt-4o-realtime-preview, it supported natural speech-to-speech conversation over a persistent WebSocket connection, with six preset voices and function calling, so a voice assistant could place an order or look up customer data mid-call.

Before this, developers chained a transcription model, a text model and a text-to-speech model, which added latency and lost tone and emphasis. Pricing was $5 and $20 per million text input and output tokens, and $100 and $200 per million audio tokens, about $0.06 per minute of audio input and $0.24 per minute of output. Early users included Healthify and Speak.

Realtime API in numbers

Values are copied from the linked source on the date shown. They are not IA Compa measurements.

Audio input price per 1M audio tokens · gpt-4o-realtime-previewas of · Source for Audio input price per 1M audio tokens: openai.com (opens in a new tab)
$100 (~$0.06/min)
Audio output price per 1M audio tokens · gpt-4o-realtime-previewas of · Source for Audio output price per 1M audio tokens: openai.com (opens in a new tab)
$200 (~$0.24/min)

Why Realtime API matters for businesses

For businesses, this was the foundation for modern AI phone agents. Receptionists, appointment setters and support lines could respond in about the time a person would, though teams still needed consent, disclosure and call-recording rules in place.

Put it to work: AI Voice Agents

Sources for this story

  1. Official
  2. Official

Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.

  • #voice-ai
  • #openai
  • #speech-to-speech
  • #api
  • #voice-agents

Want this working in your business?

We turn new AI capabilities into agents, voice receptionists, and automations you can actually use.