Back to model catalog
GPT Realtime 2
OpenAI GPT Realtime
Released 2026-05-07
ChatJSON modePrompt cachingStreamingTool calling
128K
Context window
Audio, Image, Text
Modalities
$4.00 / $24.00
Input / output per M tokens
Providers & pricing (1)
| Provider | Input $/M | Output $/M | Cached input $/M | Context |
|---|---|---|---|---|
OpenAI gpt-realtime-2 | $4.00 | $24.00 | $0.40 | 128K |
Cloptima's known static list pricing per million tokens; live/negotiated rates may differ.
Example: 1,000 requests of 2,000 input and 500 output tokens costs about $20.00 on OpenAI.
Use it through Cloptima
curl https://api.cloptima.ai/v1/ai/chat/completions \
-H "Authorization: Bearer $CLOPTIMA_PAT" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-realtime-2",
"messages": [{"role": "user", "content": "Hello"}]
}'Use the bare model name and Cloptima picks the provider for each request: the lowest estimated cost among the providers your credentials or managed credits can use, filtered by your policy and provider health.
An OpenAI-compatible endpoint with per-team cost attribution, budgets and policy controls.
How the AI gateway works