GPT Audio

OpenAI · 2 models · Text, Audio → Text, Audio

POSThttps://onerouter.com/v1/chat/completions

How to call

Pass any model ID from this series in the model field, e.g. openai/gpt-audio. The example on the right uses GPT Audio.

Overview

GPT Audio is OpenAI's family of speech-capable chat models that accept audio input and return audio or text, with tool use and structured outputs.


This series includes 2 models: openai/gpt-audio, openai/gpt-audio-mini. They share the same request shape — switching between them is a one-line change to the model field of the Chat Completions API.

Models in this series

Switch tabs to compare each model's ID, context window, inputs, outputs, and capabilities. Check the Models page for current availability and pricing before use.

IDopenai/gpt-audio
ProviderOpenAI
Context window128K
Max output16.4K
InputText, Audio
OutputText, Audio
SummaryGPT Audio is OpenAI's speech-capable chat model. It accepts audio input and returns audio (or text) output, with tool use and structured outputs.
StrengthsNatural spoken responses, audio understanding, and tool calling in one model — good for voice agents.
Supported parameters`frequency_penalty`, `logit_bias`, `logprobs`, `max_tokens`, `presence_penalty`, `response_format`, `seed`, `stop`, `structured_outputs`, `temperature`

Request headers

ParameterTypeRequiredDescription
Authorizationstring · requiredrequiredBearer token, formatted as `Bearer YOUR_API_KEY`.
Content-Typestring · requiredrequiredMust be `application/json`.

Request body

Audio-capable models are called through Chat Completions with audio content parts, or through the dedicated audio routes. Common chat fields:

ParameterTypeRequiredDescriptionValues
modelstringrequiredAudio model ID from this series (see the tabs above).
messagesarrayrequiredConversation messages. Content may include `input_audio` parts (base64 audio + format).
modalitiesarrayoptionalOutput modalities to request, e.g. `["text", "audio"]`.
audioobjectoptionalOutput audio config, e.g. `{"voice": "alloy", "format": "mp3"}`.
max_tokensintegeroptionalMaximum tokens to generate.
temperaturenumberoptionalSampling temperature.
streambooleanoptionalStream the response as SSE.
truefalse

Response fields

A non-streaming call returns a chat.completion object (see the panel on the right).

ParameterTypeRequiredDescription
idstringoptionalUnique identifier for the completion.
choicesarrayoptionalGenerated choices. `choices[0].message.content` holds the reply.
choices[].finish_reasonstringoptionalWhy generation stopped: `stop`, `length`, `tool_calls`, or `content_filter`.
usageobjectoptionalToken accounting: `prompt_tokens`, `completion_tokens`, `total_tokens`.
modelstringoptionalThe model that produced the response.

Model-dependent features

OneRouter keeps protocol compatibility and forwards model-specific parameters where supported. Features such as tool calling, structured output, reasoning, prompt caching, web search, image input, audio input, video generation, and provider-specific options depend on the selected model and upstream provider. Check the model detail page before relying on a feature in production.