OpenAI's GPT Realtime is a proprietary multimodal chat model that accepts text, image, and audio inputs and generates text or audio outputs. It is designed for conversational tasks and can call external tools for web search, code execution, and structured data extraction via its API. The model supports streaming, parallel tool calls, and extended reasoning for multi-step tasks.
Input
Output
Context
32K
Max Output
4K
Parameters
-
Input Modalities
Output Modalities
Some values come from web research and may be incorrect. Report incorrect data
None the catalog can show from stated facts or measured scores.
None the catalog can show from stated facts or measured scores.
No comparable model in the catalog yet.
Specification sources
As of Nov 11, 2026.
Answered from the facts on this page, with their sources.
GPT Realtime has a context window of 32,000 tokens and a maximum output of 4,096 tokens (source: Web research). A provider may serve less; Inferbase routes on what each provider serves.
No. GPT Realtime is not released under an open license. Use it through OpenAI's API or a provider that serves it.
OpenAI has not published a parameter count for GPT Realtime. The catalog does not estimate undisclosed sizes.
GPT Realtime accepts Text, Image and Audio and produces Text and Audio. The Capabilities card lists the API features it supports, such as function calling and structured output.
No benchmark source has published a result for GPT Realtime yet, so it is shown as unmeasured, not estimated.
Not on Inferbase Managed today. Add your own provider key or endpoint that serves GPT Realtime, and Inferbase runs it on your account with the same routing and logs.
OpenAI does not publish open weights for GPT Realtime, so it runs through an API: OpenAI's own, or Inferbase where it is served.
Start on the free plan with 5,000 requests a month, no card required.