GPT Realtime Mini

A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.

Input price: $0.66
Output price: $2.64

Context: 32K
Max output: 4K

30-day uptime: —

Released: 2025-10-07

InputTextAudioImageOutputTextAudio

This model is being retired

Requests stop working on 2027-01-20.

Providers

This model is served by the dedicated providers below. We pick the provider for each request, so provider preferences don’t apply, including a provider named in your request. Prices in USD per 1M tokens. Need a custom setup? Contact [email protected].

DedicatedInput$0.66Output$2.64Cache read$0.066Context32KMax output4KThis model is served by a dedicated set of providers. We pick the provider for each request, and you pay the same price whichever one serves it.

Included providers1

SuccessTTFT
OpenAIUnited States——

Pricing

Prices for this model. Each request is billed at these prices, whichever provider serves it.

Input: $0.66; per 1M tokens
Output: $2.64; per 1M tokens
Cache read: $0.066; per 1M tokens
Audio input: $11.00; per 1M tokens
Audio cache read: $0.33; per 1M tokens
Audio output: $22.00; per 1M tokens
Image input: $0.88; per 1M tokens
Image cache read: $0.088; per 1M tokens

Try it

Talk to the model live on your balance. Each reply is billed as it finishes, and your browser will ask for the microphone.

{
  "type": "realtime",
  "model": "openai/gpt-realtime-mini",
  "audio": {
    "input": {
      "transcription": {
        "model": "gpt-4o-mini-transcribe"
      }
    },
    "output": {
      "voice": "marin"
    }
  }
}

// EPHEMERAL_KEY: mint one on your server with POST /v1/realtime/client_secrets
const peer = new RTCPeerConnection();
const speaker = new Audio();
speaker.autoplay = true;
peer.ontrack = (event) => {
  speaker.srcObject = event.streams[0];
};

const microphone = await navigator.mediaDevices.getUserMedia({ audio: true });
peer.addTrack(microphone.getTracks()[0], microphone);
const events = peer.createDataChannel("oai-events");
events.onmessage = (event) => console.log(JSON.parse(event.data));

await peer.setLocalDescription(await peer.createOffer());
const form = new FormData();
form.set("sdp", peer.localDescription.sdp);
form.set("session", JSON.stringify({
  type: "realtime",
  model: "openai/gpt-realtime-mini",
  audio: {
    input: { transcription: { model: "gpt-4o-mini-transcribe" } },
    output: { voice: "marin" },
  },
}));
const response = await fetch("https://api.ohmygpt.com/v1/realtime/calls", {
  method: "POST",
  headers: { Authorization: "Bearer " + EPHEMERAL_KEY },
  body: form,
});
await peer.setRemoteDescription({ type: "answer", sdp: await response.text() });

Start a call to hear the model and see both sides live.

Billed per reply. Hanging up stops billing.

Similar models

Other models with the same output types at a similar price.

GPT Realtime MiniThis model

Input: $0.66
Output: $2.64

GPT Realtime 2.1 Mini

Input: $0.66
Output: $2.64

Gemini 3.8 Live

Input: $0.675
Output: $4.05