Create client secret
POST
/realtime/client_secretsCreate a Realtime client secret with an associated session configuration.
Client secrets are short-lived tokens that can be passed to a client app, such as a web frontend or mobile client, which grants access to the Realtime API without leaking your main API key. You can configure a custom TTL for each client secret.
You can also attach session configuration options to the client secret, which will be applied to any sessions created using that client secret, but these can also be overridden by the client connection.
Learn more about authentication with client secrets over WebRTC.
Returns the created client secret and the effective session object. The client secret is a string that looks like ek_1234.
Create client secret
curl -X POST https://api.openai.com/v1/realtime/client_secrets \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"expires_after": {
"anchor": "created_at",
"seconds": 600
},
"session": {
"type": "realtime",
"model": "gpt-realtime",
"instructions": "You are a friendly assistant."
}
}'
Response example
{
"value": "ek_68af296e8e408191a1120ab6383263c2",
"expires_at": 1756310470,
"session": {
"type": "realtime",
"object": "realtime.session",
"id": "sess_C9CiUVUzUzYIssh3ELY1d",
"model": "gpt-realtime",
"output_modalities": [
"audio"
],
"instructions": "You are a friendly assistant.",
"tools": [],
"tool_choice": "auto",
"max_output_tokens": "inf",
"tracing": null,
"truncation": "auto",
"prompt": null,
"expires_at": 0,
"audio": {
"input": {
"format": {
"type": "audio/pcm",
"rate": 24000
},
"transcription": null,
"noise_reduction": null,
"turn_detection": {
"type": "server_vad"
}
},
"output": {
"format": {
"type": "audio/pcm",
"rate": 24000
},
"voice": "alloy",
"speed": 1.0
}
},
"include": null
}
}
Body ParametersJSON
expires_after: optional object { anchor, seconds } Configuration for the client secret expiration. Expiration refers to the time after which
a client secret will no longer be valid for creating sessions. The session itself may
continue after that time once started. A secret can be used to create multiple sessions
until it expires.
expires_after: optional object { anchor, seconds }
Configuration for the client secret expiration. Expiration refers to the time after which a client secret will no longer be valid for creating sessions. The session itself may continue after that time once started. A secret can be used to create multiple sessions until it expires.
session: optional RealtimeSessionCreateRequest { type, audio, include, 11 more } or RealtimeTranscriptionSessionCreateRequest { type, audio, include } Session configuration to use for the client secret. Choose either a realtime
session or a transcription session.
session: optional RealtimeSessionCreateRequest { type, audio, include, 11 more } or RealtimeTranscriptionSessionCreateRequest { type, audio, include }
Session configuration to use for the client secret. Choose either a realtime session or a transcription session.
One of the following: RealtimeSessionCreateRequest { type, audio, include, 11 more } Realtime session object configuration.
type: "realtime" The type of session to create. Always realtime for the Realtime API.
audio: optional RealtimeAudioConfig { input, output } Configuration for input and output audio.
input: optional RealtimeAudioConfigInput { format, noise_reduction, transcription, turn_detection } format: optional RealtimeAudioFormats The format of the input audio.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
noise_reduction: optional { strength, type } Configuration for input audio noise reduction. This can be set to null to turn off.
Noise reduction filters audio added to the input audio buffer before it is sent to VAD and the model.
Filtering the audio can improve VAD and turn detection accuracy (reducing false positives) and model performance by improving perception of the input audio.
strength: optional number (default: 1) Input noise suppression strength. Higher values apply more suppression.
Defaults to 1 when noise reduction is first enabled. Omitting strength
from a session update preserves the current strength.
minimum0 maximum1 type: optional NoiseReductionType Type of noise reduction. near_field is for close-talking microphones such as headphones, far_field is for far-field microphones such as laptop or conference room microphones.
One of the following: "near_field" "far_field" transcription: optional AudioTranscription { delay, keywords, language, 3 more } Configuration for input audio transcription, defaults to off and can be set to null to turn off once on. Input audio transcription is not native to the model, since the model consumes audio directly. Transcription runs asynchronously through the /audio/transcriptions endpoint and should be treated as guidance of input audio content rather than precisely what the model heard. The client can optionally set the language and prompt for transcription, these offer additional guidance to the transcription service.
delay: optional "minimal" or "low" or "medium" or 2 more Controls how long the model waits before emitting transcription text.
Higher values can improve transcription accuracy at the cost of latency.
Only supported with gpt-realtime-whisper in GA Realtime sessions.
One of the following: "minimal" "low" "medium" "high" "xhigh" keywords: optional array of string Words or phrases to guide transcription of the input audio. Supported by gpt-transcribe and gpt-live-transcribe.
language: optional string The language of the input audio. Supplying the input language in
ISO-639-1 (e.g. en) format
will improve accuracy and latency.
languages: optional array of string Possible languages of the input audio, in ISO-639-1 format. Supported by gpt-transcribe and gpt-live-transcribe.
minItems1 model: optional string or "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model to use for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper. Use gpt-4o-transcribe-diarize when you need diarization with speaker labels.
One of the following: string "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model to use for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper. Use gpt-4o-transcribe-diarize when you need diarization with speaker labels.
One of the following: "gpt-transcribe" "gpt-live-transcribe" "gpt-realtime-whisper" "whisper-1" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe-2025-12-15" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe-diarize" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
prompt: optional string An optional text to guide the model’s style or continue a previous audio
segment.
For whisper-1, the prompt is a list of keywords.
For gpt-4o-transcribe models (excluding gpt-4o-transcribe-diarize), the prompt is a free text string, for example “expect words related to technology”.
Prompt is not supported with gpt-realtime-whisper in GA Realtime sessions.
turn_detection: optional RealtimeAudioInputTurnDetection or null Configuration for turn detection, ether Server VAD or Semantic VAD. This can be set to null to turn off, in which case the client must manually trigger model response.
Server VAD means that the model will detect the start and end of speech based on audio volume and respond at the end of user speech.
Semantic VAD is more advanced and uses a turn detection model (in conjunction with VAD) to semantically estimate whether the user has finished speaking, then dynamically sets a timeout based on this probability. For example, if user audio trails off with “uhhm”, the model will score a low probability of turn end and wait longer for the user to continue speaking. This can be useful for more natural conversations, but may have a higher latency.
For gpt-realtime-whisper transcription sessions, turn detection must be
set to null; VAD is not supported.
One of the following: ServerVad { type, create_response, idle_timeout_ms, 4 more } Server-side voice activity detection (VAD) which flips on when user speech is detected and off after a period of silence.
type: "server_vad" (default: "server_vad") Type of turn detection, server_vad to turn on simple Server VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs. If interrupt_response is set to false this may fail to create a response if the model is already responding.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
idle_timeout_ms: optional integer or null Optional timeout after which a model response will be triggered automatically. This is
useful for situations in which a long pause from the user is unexpected, such as a phone
call. The model will effectively prompt the user to continue the conversation based
on the current context.
The timeout value will be applied after the last model response’s audio has finished playing,
i.e. it’s set to the response.done time plus audio playback duration.
An input_audio_buffer.timeout_triggered event (plus events
associated with the Response) will be emitted when the timeout is reached.
Idle timeout is currently only supported for server_vad mode.
minimum5000 maximum30000 interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt (cancel) any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs. If true then the response will be cancelled, otherwise it will continue until complete.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
prefix_padding_ms: optional integer Used only for server_vad mode. Amount of audio to include before the VAD detected speech (in
milliseconds). Defaults to 300ms.
silence_duration_ms: optional integer Used only for server_vad mode. Duration of silence to detect speech stop (in milliseconds). Defaults
to 500ms. With shorter values the model will respond more quickly,
but may jump in on short pauses from the user.
threshold: optional number Used only for server_vad mode. Activation threshold for VAD (0.0 to 1.0), this defaults to 0.5. A
higher threshold will require louder audio to activate the model, and
thus might perform better in noisy environments.
SemanticVad { type, create_response, eagerness, interrupt_response } Server-side semantic turn detection which uses a model to determine when the user has finished speaking.
type: "semantic_vad" Type of turn detection, semantic_vad to turn on Semantic VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs.
eagerness: optional "low" or "medium" or "high" or "auto" (default: "auto") Used only for semantic_vad mode. The eagerness of the model to respond. low will wait longer for the user to continue speaking, high will respond more quickly. auto is the default and is equivalent to medium. low, medium, and high have max timeouts of 8s, 4s, and 2s respectively.
One of the following: "low" "medium" "high" "auto" interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs.
output: optional RealtimeAudioConfigOutput { format, speed, voice } format: optional RealtimeAudioFormats The format of the output audio.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
speed: optional number (default: 1) The speed of the model’s spoken response as a multiple of the original speed.
1.0 is the default speed. 0.25 is the minimum speed. 1.5 is the maximum speed. This value can only be changed in between model turns, not while a response is in progress.
This parameter is a post-processing adjustment to the audio after it is generated, it’s
also possible to prompt the model to speak faster or slower.
minimum0.25 maximum1.5 voice: optional string or "alloy" or "ash" or "ballad" or 7 more or ID (default: "alloy") The voice the model uses to respond. Supported built-in voices are
alloy, ash, ballad, coral, echo, sage, shimmer, verse,
marin, and cedar. You may also provide a custom voice object with
an id, for example { "id": "voice_1234" }. Voice cannot be changed
during the session once the model has responded with audio at least once.
Custom voices must be created from audio samples. Voices created from text prompts are supported only in Live.
We recommend marin and cedar for best quality.
One of the following: string "alloy" or "ash" or "ballad" or 7 more One of the following: "alloy" "ash" "ballad" "coral" "echo" "sage" "shimmer" "verse" "marin" "cedar" ID { id } Custom voice reference.
id: string The custom voice ID, e.g. voice_1234.
include: optional array of "item.input_audio_transcription.logprobs" Additional fields to include in server outputs.
item.input_audio_transcription.logprobs: Include logprobs for input audio transcription.
instructions: optional string The default system instructions (i.e. system message) prepended to model calls. This field allows the client to guide the model on desired responses. The model can be instructed on response content and format, (e.g. “be extremely succinct”, “act friendly”, “here are examples of good responses”) and on audio behavior (e.g. “talk quickly”, “inject emotion into your voice”, “laugh frequently”). The instructions are not guaranteed to be followed by the model, but they provide guidance to the model on the desired behavior.
Note that the server sets default instructions which will be used if this field is not set and are visible in the session.created event at the start of the session.
max_output_tokens: optional number or "inf" Maximum number of output tokens for a single assistant response,
inclusive of tool calls. Provide an integer between 1 and 4096 to
limit output tokens, or inf for the maximum available tokens for a
given model. Defaults to inf.
One of the following: number "inf" model: optional string or "gpt-realtime" or "gpt-realtime-1.5" or "gpt-realtime-2" or 16 more The Realtime model used for this session.
One of the following: string "gpt-realtime" or "gpt-realtime-1.5" or "gpt-realtime-2" or 16 more The Realtime model used for this session.
One of the following: "gpt-realtime-1.5" "gpt-realtime-2" "gpt-realtime-2.1" "gpt-realtime-2.1-mini" "gpt-4o-realtime-preview" "gpt-4o-realtime-preview-2024-10-01" "gpt-4o-realtime-preview-2024-12-17" "gpt-4o-realtime-preview-2025-06-03" "gpt-4o-mini-realtime-preview" "gpt-4o-mini-realtime-preview-2024-12-17" "gpt-realtime-mini-2025-10-06" "gpt-audio-1.5" "gpt-audio-mini-2025-10-06" "gpt-realtime" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-2025-08-28" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-mini" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-mini-2025-12-15" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-audio-mini" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-audio-mini-2025-12-15" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
output_modalities: optional array of "text" or "audio" (default: ["audio"]) The set of modalities the model can respond with. It defaults to ["audio"], indicating
that the model will respond with audio plus a transcript. ["text"] can be used to make
the model respond with text only. It is not possible to request both text and audio at the same time.
One of the following: "text" "audio" parallel_tool_calls: optional boolean Whether the model may call multiple tools in parallel. Only supported by
reasoning Realtime models such as gpt-realtime-2.
prompt: optional ResponsePrompt or null Reference to a prompt template and its variables.
Learn more.
id: string The unique identifier of the prompt template to use.
variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null Optional map of values to substitute in for variables in your
prompt. The substitution values can either be strings, or other
Response input types like images or files.
One of the following: string ResponseInputText { text, type, prompt_cache_breakpoint } A text input to the model.
text: string The text input to the model.
type: "input_text" (default: "input_text") The type of the input item. Always input_text.
prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" ResponseInputImage { detail, type, file_id, 2 more } An image input to the model. Learn about image inputs.
detail: ImageDetail The detail level of the image to be sent to the model. One of high, low, auto, or original. Defaults to auto.
One of the following: "low" "high" "auto" "original" type: "input_image" (default: "input_image") The type of the input item. Always input_image.
file_id: optional string or null The ID of the file to be sent to the model.
image_url: optional string or null The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL.
formaturi prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" ResponseInputFile { type, detail, file_data, 4 more } A file input to the model.
type: "input_file" (default: "input_file") The type of the input item. Always input_file.
detail: optional "auto" or "low" or "high" The detail level of the file to be sent to the model. Use auto to let the system select the detail level; for GPT-5.6 and later models, auto uses high-quality rendering, which may increase input token usage. Use low for lower-cost rendering, or high to render the file at higher quality. Defaults to auto.
One of the following: "auto" "low" "high" file_data: optional string The content of the file to be sent to the model.
file_id: optional string or null The ID of the file to be sent to the model.
file_url: optional string The URL of the file to be sent to the model.
formaturi filename: optional string The name of the file to be sent to the model.
prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" version: optional string or null Optional version of the prompt template.
reasoning: optional RealtimeReasoning { effort } Configuration for reasoning-capable Realtime models such as gpt-realtime-2.
effort: optional RealtimeReasoningEffort (default: "low") Constrains effort on reasoning for reasoning-capable Realtime models such as
gpt-realtime-2.
One of the following: "minimal" "low" "medium" "high" "xhigh" tool_choice: optional RealtimeToolChoiceConfig (default: "auto") How the model chooses tools. Provide one of the string modes or force a specific
function/MCP tool.
One of the following: ToolChoiceOptions = "none" or "auto" or "required" Controls which (if any) tool is called by the model.
none means the model will not call any tool and instead generates a message.
auto means the model can pick between generating a message or calling one or
more tools.
required means the model must call one or more tools.
One of the following: "none" "auto" "required" ToolChoiceFunction { name, type } Use this option to force the model to call a specific function.
name: string The name of the function to call.
type: "function" For function calling, the type is always function.
ToolChoiceMcp { server_label, type, name } Use this option to force the model to call a specific tool on a remote MCP server.
server_label: string The label of the MCP server to use.
type: "mcp" For MCP tools, the type is always mcp.
name: optional string or null The name of the tool to call on the server.
tools: optional RealtimeToolsConfig { , McpTool } Tools available to the model.
One of the following: RealtimeFunctionTool { description, name, parameters, type } description: optional string The description of the function, including guidance on when and how
to call it, and guidance about what to tell the user when calling
(if anything).
name: optional string The name of the function.
parameters: optional unknown Parameters of the function in JSON Schema.
type: optional "function" The type of the tool, i.e. function.
McpTool { server_label, type, allowed_callers, 9 more } Give the model access to additional tools via remote Model Context Protocol
(MCP) servers. Learn more about MCP.
server_label: string A label for this MCP server, used to identify it in tool calls.
type: "mcp" The type of the MCP tool. Always mcp.
allowed_callers: optional array of "direct" or "programmatic" or null The tool invocation context(s).
minItems1 One of the following: "direct" "programmatic" allowed_tools: optional array of string or McpToolFilter or null List of allowed tool names or a filter object.
One of the following: McpAllowedTools = array of string A string array of allowed tool names
McpToolFilter { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
authorization: optional string An OAuth access token that can be used with a remote MCP server, either
with a custom MCP server URL or a service connector. Your application
must handle the OAuth authorization flow and provide the token here.
connector_id: optional "connector_dropbox" or "connector_gmail" or "connector_googlecalendar" or 5 more Deprecated.
Identifier for service connectors, like those available in ChatGPT. One of
server_url, connector_id, or tunnel_id must be provided. Learn more
about service connectors here.
This field is deprecated for models released after September 1, 2026.
Use server_url to connect to a remote MCP server, or tunnel_id to
connect through a Secure MCP Tunnel.
Currently supported connector_id values are:
Dropbox: connector_dropbox
Gmail: connector_gmail
Google Calendar: connector_googlecalendar
Google Drive: connector_googledrive
Microsoft Teams: connector_microsoftteams
Outlook Calendar: connector_outlookcalendar
Outlook Email: connector_outlookemail
SharePoint: connector_sharepoint
One of the following: "connector_dropbox" "connector_gmail" "connector_googlecalendar" "connector_googledrive" "connector_microsoftteams" "connector_outlookcalendar" "connector_outlookemail" "connector_sharepoint" defer_loading: optional boolean Whether this MCP tool is deferred and discovered via tool search.
headers: optional map[string] or null Optional HTTP headers to send to the MCP server. Use for authentication
or other purposes.
require_approval: optional McpToolApprovalFilter or "always" or "never" or null (default: "always") Specify which of the MCP server’s tools require approval.
One of the following: McpToolApprovalFilter { always, never } Specify which of the MCP server’s tools require approval. Can be
always, never, or a filter object associated with tools
that require approval.
always: optional { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
never: optional { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
McpToolApprovalSetting = "always" or "never" Specify a single approval policy for all tools. One of always or
never. When set to always, all tools will require approval. When
set to never, all tools will not require approval.
One of the following: "always" "never" server_description: optional string Optional description of the MCP server, used to provide more context.
server_url: optional string The URL for the MCP server. One of server_url, connector_id, or
tunnel_id must be provided.
formaturi tunnel_id: optional string The Secure MCP Tunnel ID to use instead of a direct server URL. One of
server_url, connector_id, or tunnel_id must be provided.
pattern^tunnel_[a-z0-9]{32}$ tracing: optional RealtimeTracingConfig or null Realtime API can write session traces to the Traces Dashboard. Set to null to disable tracing. Once
tracing is enabled for a session, the configuration cannot be modified.
auto will create a trace for the session with default values for the
workflow name, group id, and metadata.
One of the following: Auto = "auto" Enables tracing and sets default values for tracing configuration options. Always auto.
TracingConfiguration { group_id, metadata, workflow_name } Granular configuration for tracing.
group_id: optional string The group id to attach to this trace to enable filtering and
grouping in the Traces Dashboard.
metadata: optional unknown The arbitrary metadata to attach to this trace to enable
filtering in the Traces Dashboard.
workflow_name: optional string The name of the workflow to attach to this trace. This is used to
name the trace in the Traces Dashboard.
truncation: optional RealtimeTruncation When the number of tokens in a conversation exceeds the model’s input token limit, the conversation be truncated, meaning messages (starting from the oldest) will not be included in the model’s context. A 32k context model with 4,096 max output tokens can only include 28,224 tokens in the context before truncation occurs.
Clients can configure truncation behavior to truncate with a lower max token limit, which is an effective way to control token usage and cost.
Truncation will reduce the number of cached tokens on the next turn (busting the cache), since messages are dropped from the beginning of the context. However, clients can also configure truncation to retain messages up to a fraction of the maximum context size, which will reduce the need for future truncations and thus improve the cache rate.
Truncation can be disabled entirely, which means the server will never truncate but would instead return an error if the conversation exceeds the model’s input token limit.
One of the following: "auto" or "disabled" The truncation strategy to use for the session. auto is the default truncation strategy. disabled will disable truncation and emit errors when the conversation exceeds the input token limit.
One of the following: "auto" "disabled" RetentionRatioTruncation { retention_ratio, type, token_limits } Retain a fraction of the conversation tokens when the conversation exceeds the input token limit. This allows you to amortize truncations across multiple turns, which can help improve cached token usage.
retention_ratio: number Fraction of post-instruction conversation tokens to retain (0.0 - 1.0) when the conversation exceeds the input token limit. Setting this to 0.8 means that messages will be dropped until 80% of the maximum allowed tokens are used. This helps reduce the frequency of truncations and improve cache rates.
minimum0 maximum1 type: "retention_ratio" Use retention ratio truncation.
token_limits: optional { post_instructions } Optional custom token limits for this truncation strategy. If not provided, the model’s default token limits will be used.
post_instructions: optional integer Maximum tokens allowed in the conversation after instructions (which including tool definitions). For example, setting this to 5,000 would mean that truncation would occur when the conversation exceeds 5,000 tokens after instructions. This cannot be higher than the model’s context window size minus the maximum output tokens.
minimum0 RealtimeTranscriptionSessionCreateRequest { type, audio, include } Realtime transcription session object configuration.
type: "transcription" The type of session to create. Always transcription for transcription sessions.
audio: optional RealtimeTranscriptionSessionAudio { input } Configuration for input and output audio.
input: optional RealtimeTranscriptionSessionAudioInput { format, noise_reduction, transcription, turn_detection } format: optional RealtimeAudioFormats The PCM audio format. Only a 24kHz sample rate is supported.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
noise_reduction: optional { strength, type } Configuration for input audio noise reduction. This can be set to null to turn off.
Noise reduction filters audio added to the input audio buffer before it is sent to VAD and the model.
Filtering the audio can improve VAD and turn detection accuracy (reducing false positives) and model performance by improving perception of the input audio.
strength: optional number (default: 1) Input noise suppression strength. Higher values apply more suppression.
Defaults to 1 when noise reduction is first enabled. Omitting strength
from a session update preserves the current strength.
minimum0 maximum1 type: optional NoiseReductionType Type of noise reduction. near_field is for close-talking microphones such as headphones, far_field is for far-field microphones such as laptop or conference room microphones.
One of the following: "near_field" "far_field" transcription: optional AudioTranscription { delay, keywords, language, 3 more } Configuration for input audio transcription, defaults to off and can be set to null to turn off once on. Input audio transcription is not native to the model, since the model consumes audio directly. Transcription runs asynchronously through the /audio/transcriptions endpoint and should be treated as guidance of input audio content rather than precisely what the model heard. The client can optionally set the language and prompt for transcription, these offer additional guidance to the transcription service.
delay: optional "minimal" or "low" or "medium" or 2 more Controls how long the model waits before emitting transcription text.
Higher values can improve transcription accuracy at the cost of latency.
Only supported with gpt-realtime-whisper in GA Realtime sessions.
One of the following: "minimal" "low" "medium" "high" "xhigh" keywords: optional array of string Words or phrases to guide transcription of the input audio. Supported by gpt-transcribe and gpt-live-transcribe.
language: optional string The language of the input audio. Supplying the input language in
ISO-639-1 (e.g. en) format
will improve accuracy and latency.
languages: optional array of string Possible languages of the input audio, in ISO-639-1 format. Supported by gpt-transcribe and gpt-live-transcribe.
minItems1 model: optional string or "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model to use for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper. Use gpt-4o-transcribe-diarize when you need diarization with speaker labels.
One of the following: string "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model to use for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper. Use gpt-4o-transcribe-diarize when you need diarization with speaker labels.
One of the following: "gpt-transcribe" "gpt-live-transcribe" "gpt-realtime-whisper" "whisper-1" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe-2025-12-15" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe-diarize" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
prompt: optional string An optional text to guide the model’s style or continue a previous audio
segment.
For whisper-1, the prompt is a list of keywords.
For gpt-4o-transcribe models (excluding gpt-4o-transcribe-diarize), the prompt is a free text string, for example “expect words related to technology”.
Prompt is not supported with gpt-realtime-whisper in GA Realtime sessions.
turn_detection: optional RealtimeTranscriptionSessionAudioInputTurnDetection or null Configuration for turn detection, ether Server VAD or Semantic VAD. This can be set to null to turn off, in which case the client must manually trigger model response.
Server VAD means that the model will detect the start and end of speech based on audio volume and respond at the end of user speech.
Semantic VAD is more advanced and uses a turn detection model (in conjunction with VAD) to semantically estimate whether the user has finished speaking, then dynamically sets a timeout based on this probability. For example, if user audio trails off with “uhhm”, the model will score a low probability of turn end and wait longer for the user to continue speaking. This can be useful for more natural conversations, but may have a higher latency.
For gpt-realtime-whisper transcription sessions, turn detection must be
set to null; VAD is not supported.
One of the following: ServerVad { type, create_response, idle_timeout_ms, 4 more } Server-side voice activity detection (VAD) which flips on when user speech is detected and off after a period of silence.
type: "server_vad" (default: "server_vad") Type of turn detection, server_vad to turn on simple Server VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs. If interrupt_response is set to false this may fail to create a response if the model is already responding.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
idle_timeout_ms: optional integer or null Optional timeout after which a model response will be triggered automatically. This is
useful for situations in which a long pause from the user is unexpected, such as a phone
call. The model will effectively prompt the user to continue the conversation based
on the current context.
The timeout value will be applied after the last model response’s audio has finished playing,
i.e. it’s set to the response.done time plus audio playback duration.
An input_audio_buffer.timeout_triggered event (plus events
associated with the Response) will be emitted when the timeout is reached.
Idle timeout is currently only supported for server_vad mode.
minimum5000 maximum30000 interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt (cancel) any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs. If true then the response will be cancelled, otherwise it will continue until complete.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
prefix_padding_ms: optional integer Used only for server_vad mode. Amount of audio to include before the VAD detected speech (in
milliseconds). Defaults to 300ms.
silence_duration_ms: optional integer Used only for server_vad mode. Duration of silence to detect speech stop (in milliseconds). Defaults
to 500ms. With shorter values the model will respond more quickly,
but may jump in on short pauses from the user.
threshold: optional number Used only for server_vad mode. Activation threshold for VAD (0.0 to 1.0), this defaults to 0.5. A
higher threshold will require louder audio to activate the model, and
thus might perform better in noisy environments.
SemanticVad { type, create_response, eagerness, interrupt_response } Server-side semantic turn detection which uses a model to determine when the user has finished speaking.
type: "semantic_vad" Type of turn detection, semantic_vad to turn on Semantic VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs.
eagerness: optional "low" or "medium" or "high" or "auto" (default: "auto") Used only for semantic_vad mode. The eagerness of the model to respond. low will wait longer for the user to continue speaking, high will respond more quickly. auto is the default and is equivalent to medium. low, medium, and high have max timeouts of 8s, 4s, and 2s respectively.
One of the following: "low" "medium" "high" "auto" interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs.
include: optional array of "item.input_audio_transcription.logprobs" Additional fields to include in server outputs.
item.input_audio_transcription.logprobs: Include logprobs for input audio transcription.
Returns
expires_at: integer
Expiration timestamp for the client secret, in seconds since epoch.
formatunixtime
session: RealtimeSessionCreateResponse { id, object, type, 13 more } or RealtimeTranscriptionSessionCreateResponse { id, object, type, 3 more } The session configuration for either a realtime or transcription session.
session: RealtimeSessionCreateResponse { id, object, type, 13 more } or RealtimeTranscriptionSessionCreateResponse { id, object, type, 3 more }
The session configuration for either a realtime or transcription session.
One of the following: RealtimeSessionCreateResponse { id, object, type, 13 more } A Realtime session configuration object.
id: string Unique identifier for the session that looks like sess_1234567890abcdef.
object: "realtime.session" The object type. Always realtime.session.
type: "realtime" The type of session to create. Always realtime for the Realtime API.
audio: optional { input, output } Configuration for input and output audio.
input: optional { format, noise_reduction, transcription, turn_detection } format: optional RealtimeAudioFormats The format of the input audio.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
noise_reduction: optional object or null (default: null) Configuration for input audio noise reduction. This can be set to null to turn off.
Noise reduction filters audio added to the input audio buffer before it is sent to VAD and the model.
Filtering the audio can improve VAD and turn detection accuracy (reducing false positives) and model performance by improving perception of the input audio.
strength: optional number (default: 1) Input noise suppression strength. Higher values apply more suppression.
Defaults to 1 when noise reduction is first enabled. Omitting strength
from a session update preserves the current strength.
minimum0 maximum1 type: optional NoiseReductionType Type of noise reduction. near_field is for close-talking microphones such as headphones, far_field is for far-field microphones such as laptop or conference room microphones.
One of the following: "near_field" "far_field" transcription: optional object or null Configuration for input audio transcription, defaults to off and can be set to null to turn off once on. Input audio transcription is not native to the model, since the model consumes audio directly. Transcription runs asynchronously through the /audio/transcriptions endpoint and should be treated as guidance of input audio content rather than precisely what the model heard. The client can optionally set the language and prompt for transcription, these offer additional guidance to the transcription service.
keywords: optional array of string Words or phrases configured to guide transcription of the input audio.
language: optional string or null The language of the input audio.
languages: optional array of string The possible input audio languages configured for transcription, in ISO-639-1 format.
minItems1 model: optional string or "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model used for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper.
One of the following: string "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model used for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper.
One of the following: "gpt-transcribe" "gpt-live-transcribe" "gpt-realtime-whisper" "whisper-1" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe-2025-12-15" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe-diarize" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
prompt: optional string The prompt configured for input audio transcription, when present.
turn_detection: optional ServerVad or SemanticVad or null Configuration for turn detection, ether Server VAD or Semantic VAD. This can be set to null to turn off, in which case the client must manually trigger model response.
Server VAD means that the model will detect the start and end of speech based on audio volume and respond at the end of user speech.
Semantic VAD is more advanced and uses a turn detection model (in conjunction with VAD) to semantically estimate whether the user has finished speaking, then dynamically sets a timeout based on this probability. For example, if user audio trails off with “uhhm”, the model will score a low probability of turn end and wait longer for the user to continue speaking. This can be useful for more natural conversations, but may have a higher latency.
For gpt-realtime-whisper transcription sessions, turn detection must be
set to null; VAD is not supported.
One of the following: ServerVad { type, create_response, idle_timeout_ms, 4 more } Server-side voice activity detection (VAD) which flips on when user speech is detected and off after a period of silence.
type: "server_vad" (default: "server_vad") Type of turn detection, server_vad to turn on simple Server VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs. If interrupt_response is set to false this may fail to create a response if the model is already responding.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
idle_timeout_ms: optional integer or null Optional timeout after which a model response will be triggered automatically. This is
useful for situations in which a long pause from the user is unexpected, such as a phone
call. The model will effectively prompt the user to continue the conversation based
on the current context.
The timeout value will be applied after the last model response’s audio has finished playing,
i.e. it’s set to the response.done time plus audio playback duration.
An input_audio_buffer.timeout_triggered event (plus events
associated with the Response) will be emitted when the timeout is reached.
Idle timeout is currently only supported for server_vad mode.
minimum5000 maximum30000 interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt (cancel) any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs. If true then the response will be cancelled, otherwise it will continue until complete.
If both create_response and interrupt_response are set to false, the model will never respond automatically but VAD events will still be emitted.
prefix_padding_ms: optional integer Used only for server_vad mode. Amount of audio to include before the VAD detected speech (in
milliseconds). Defaults to 300ms.
silence_duration_ms: optional integer Used only for server_vad mode. Duration of silence to detect speech stop (in milliseconds). Defaults
to 500ms. With shorter values the model will respond more quickly,
but may jump in on short pauses from the user.
threshold: optional number Used only for server_vad mode. Activation threshold for VAD (0.0 to 1.0), this defaults to 0.5. A
higher threshold will require louder audio to activate the model, and
thus might perform better in noisy environments.
SemanticVad { type, create_response, eagerness, interrupt_response } Server-side semantic turn detection which uses a model to determine when the user has finished speaking.
type: "semantic_vad" Type of turn detection, semantic_vad to turn on Semantic VAD.
create_response: optional boolean (default: true) Whether or not to automatically generate a response when a VAD stop event occurs.
eagerness: optional "low" or "medium" or "high" or "auto" (default: "auto") Used only for semantic_vad mode. The eagerness of the model to respond. low will wait longer for the user to continue speaking, high will respond more quickly. auto is the default and is equivalent to medium. low, medium, and high have max timeouts of 8s, 4s, and 2s respectively.
One of the following: "low" "medium" "high" "auto" interrupt_response: optional boolean (default: true) Whether or not to automatically interrupt any ongoing response with output to the default
conversation (i.e. conversation of auto) when a VAD start event occurs.
output: optional { format, speed, voice } format: optional RealtimeAudioFormats The format of the output audio.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
speed: optional number (default: 1) The speed of the model’s spoken response as a multiple of the original speed.
1.0 is the default speed. 0.25 is the minimum speed. 1.5 is the maximum speed. This value can only be changed in between model turns, not while a response is in progress.
This parameter is a post-processing adjustment to the audio after it is generated, it’s
also possible to prompt the model to speak faster or slower.
minimum0.25 maximum1.5 voice: optional string or "alloy" or "ash" or "ballad" or 7 more (default: "alloy") The voice the model uses to respond. Voice cannot be changed during the
session once the model has responded with audio at least once. Current
voice options are alloy, ash, ballad, coral, echo, sage,
shimmer, verse, marin, and cedar. We recommend marin and cedar for
best quality.
One of the following: string "alloy" or "ash" or "ballad" or 7 more The voice the model uses to respond. Voice cannot be changed during the
session once the model has responded with audio at least once. Current
voice options are alloy, ash, ballad, coral, echo, sage,
shimmer, verse, marin, and cedar. We recommend marin and cedar for
best quality.
One of the following: "alloy" "ash" "ballad" "coral" "echo" "sage" "shimmer" "verse" "marin" "cedar" expires_at: optional integer Expiration timestamp for the session, in seconds since epoch.
formatunixtime include: optional array of "item.input_audio_transcription.logprobs" or null Additional fields to include in server outputs.
item.input_audio_transcription.logprobs: Include logprobs for input audio transcription.
instructions: optional string The default system instructions (i.e. system message) prepended to model calls. This field allows the client to guide the model on desired responses. The model can be instructed on response content and format, (e.g. “be extremely succinct”, “act friendly”, “here are examples of good responses”) and on audio behavior (e.g. “talk quickly”, “inject emotion into your voice”, “laugh frequently”). The instructions are not guaranteed to be followed by the model, but they provide guidance to the model on the desired behavior.
Note that the server sets default instructions which will be used if this field is not set and are visible in the session.created event at the start of the session.
max_output_tokens: optional number or "inf" Maximum number of output tokens for a single assistant response,
inclusive of tool calls. Provide an integer between 1 and 4096 to
limit output tokens, or inf for the maximum available tokens for a
given model. Defaults to inf.
One of the following: number "inf" model: optional string or "gpt-realtime" or "gpt-realtime-1.5" or "gpt-realtime-2" or 16 more The Realtime model used for this session.
One of the following: string "gpt-realtime" or "gpt-realtime-1.5" or "gpt-realtime-2" or 16 more The Realtime model used for this session.
One of the following: "gpt-realtime-1.5" "gpt-realtime-2" "gpt-realtime-2.1" "gpt-realtime-2.1-mini" "gpt-4o-realtime-preview" "gpt-4o-realtime-preview-2024-10-01" "gpt-4o-realtime-preview-2024-12-17" "gpt-4o-realtime-preview-2025-06-03" "gpt-4o-mini-realtime-preview" "gpt-4o-mini-realtime-preview-2024-12-17" "gpt-realtime-mini-2025-10-06" "gpt-audio-1.5" "gpt-audio-mini-2025-10-06" "gpt-realtime" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-2025-08-28" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-mini" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-realtime-mini-2025-12-15" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-audio-mini" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-audio-mini-2025-12-15" Deprecated. Announced shutdown date: 2027-01-20. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
output_modalities: optional array of "text" or "audio" (default: ["audio"]) The set of modalities the model can respond with. It defaults to ["audio"], indicating
that the model will respond with audio plus a transcript. ["text"] can be used to make
the model respond with text only. It is not possible to request both text and audio at the same time.
One of the following: "text" "audio" prompt: optional ResponsePrompt or null Reference to a prompt template and its variables.
Learn more.
id: string The unique identifier of the prompt template to use.
variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null Optional map of values to substitute in for variables in your
prompt. The substitution values can either be strings, or other
Response input types like images or files.
One of the following: string ResponseInputText { text, type, prompt_cache_breakpoint } A text input to the model.
text: string The text input to the model.
type: "input_text" (default: "input_text") The type of the input item. Always input_text.
prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" ResponseInputImage { detail, type, file_id, 2 more } An image input to the model. Learn about image inputs.
detail: ImageDetail The detail level of the image to be sent to the model. One of high, low, auto, or original. Defaults to auto.
One of the following: "low" "high" "auto" "original" type: "input_image" (default: "input_image") The type of the input item. Always input_image.
file_id: optional string or null The ID of the file to be sent to the model.
image_url: optional string or null The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL.
formaturi prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" ResponseInputFile { type, detail, file_data, 4 more } A file input to the model.
type: "input_file" (default: "input_file") The type of the input item. Always input_file.
detail: optional "auto" or "low" or "high" The detail level of the file to be sent to the model. Use auto to let the system select the detail level; for GPT-5.6 and later models, auto uses high-quality rendering, which may increase input token usage. Use low for lower-cost rendering, or high to render the file at higher quality. Defaults to auto.
One of the following: "auto" "low" "high" file_data: optional string The content of the file to be sent to the model.
file_id: optional string or null The ID of the file to be sent to the model.
file_url: optional string The URL of the file to be sent to the model.
formaturi filename: optional string The name of the file to be sent to the model.
prompt_cache_breakpoint: optional { mode, ttl } Marks the exact end of a reusable prompt prefix. When ttl is omitted, the breakpoint inherits the request’s prompt_cache_options.ttl; the boundary is not rounded to a token block.
mode: "explicit" (default: "explicit") The breakpoint mode. Always explicit.
ttl: optional "30m" or "12h" The minimum lifetime of this cache breakpoint. Defaults to the request TTL.
One of the following: "30m" "12h" version: optional string or null Optional version of the prompt template.
reasoning: optional RealtimeReasoning { effort } Configuration for reasoning-capable Realtime models such as gpt-realtime-2.
effort: optional RealtimeReasoningEffort (default: "low") Constrains effort on reasoning for reasoning-capable Realtime models such as
gpt-realtime-2.
One of the following: "minimal" "low" "medium" "high" "xhigh" tool_choice: optional ToolChoiceOptions or ToolChoiceFunction or ToolChoiceMcp (default: "auto") How the model chooses tools. Provide one of the string modes or force a specific
function/MCP tool.
One of the following: ToolChoiceOptions = "none" or "auto" or "required" Controls which (if any) tool is called by the model.
none means the model will not call any tool and instead generates a message.
auto means the model can pick between generating a message or calling one or
more tools.
required means the model must call one or more tools.
One of the following: "none" "auto" "required" ToolChoiceFunction { name, type } Use this option to force the model to call a specific function.
name: string The name of the function to call.
type: "function" For function calling, the type is always function.
ToolChoiceMcp { server_label, type, name } Use this option to force the model to call a specific tool on a remote MCP server.
server_label: string The label of the MCP server to use.
type: "mcp" For MCP tools, the type is always mcp.
name: optional string or null The name of the tool to call on the server.
tools: optional array of RealtimeFunctionTool or McpTool Tools available to the model.
One of the following: RealtimeFunctionTool { description, name, parameters, type } description: optional string The description of the function, including guidance on when and how
to call it, and guidance about what to tell the user when calling
(if anything).
name: optional string The name of the function.
parameters: optional unknown Parameters of the function in JSON Schema.
type: optional "function" The type of the tool, i.e. function.
McpTool { server_label, type, allowed_callers, 9 more } Give the model access to additional tools via remote Model Context Protocol
(MCP) servers. Learn more about MCP.
server_label: string A label for this MCP server, used to identify it in tool calls.
type: "mcp" The type of the MCP tool. Always mcp.
allowed_callers: optional array of "direct" or "programmatic" or null The tool invocation context(s).
minItems1 One of the following: "direct" "programmatic" allowed_tools: optional array of string or McpToolFilter or null List of allowed tool names or a filter object.
One of the following: McpAllowedTools = array of string A string array of allowed tool names
McpToolFilter { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
authorization: optional string An OAuth access token that can be used with a remote MCP server, either
with a custom MCP server URL or a service connector. Your application
must handle the OAuth authorization flow and provide the token here.
connector_id: optional "connector_dropbox" or "connector_gmail" or "connector_googlecalendar" or 5 more Deprecated.
Identifier for service connectors, like those available in ChatGPT. One of
server_url, connector_id, or tunnel_id must be provided. Learn more
about service connectors here.
This field is deprecated for models released after September 1, 2026.
Use server_url to connect to a remote MCP server, or tunnel_id to
connect through a Secure MCP Tunnel.
Currently supported connector_id values are:
Dropbox: connector_dropbox
Gmail: connector_gmail
Google Calendar: connector_googlecalendar
Google Drive: connector_googledrive
Microsoft Teams: connector_microsoftteams
Outlook Calendar: connector_outlookcalendar
Outlook Email: connector_outlookemail
SharePoint: connector_sharepoint
One of the following: "connector_dropbox" "connector_gmail" "connector_googlecalendar" "connector_googledrive" "connector_microsoftteams" "connector_outlookcalendar" "connector_outlookemail" "connector_sharepoint" defer_loading: optional boolean Whether this MCP tool is deferred and discovered via tool search.
headers: optional map[string] or null Optional HTTP headers to send to the MCP server. Use for authentication
or other purposes.
require_approval: optional McpToolApprovalFilter or "always" or "never" or null (default: "always") Specify which of the MCP server’s tools require approval.
One of the following: McpToolApprovalFilter { always, never } Specify which of the MCP server’s tools require approval. Can be
always, never, or a filter object associated with tools
that require approval.
always: optional { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
never: optional { read_only, tool_names } A filter object to specify which tools are allowed.
read_only: optional boolean Indicates whether or not a tool modifies data or is read-only. If an
MCP server is annotated with readOnlyHint,
it will match this filter.
tool_names: optional array of string List of allowed tool names.
McpToolApprovalSetting = "always" or "never" Specify a single approval policy for all tools. One of always or
never. When set to always, all tools will require approval. When
set to never, all tools will not require approval.
One of the following: "always" "never" server_description: optional string Optional description of the MCP server, used to provide more context.
server_url: optional string The URL for the MCP server. One of server_url, connector_id, or
tunnel_id must be provided.
formaturi tunnel_id: optional string The Secure MCP Tunnel ID to use instead of a direct server URL. One of
server_url, connector_id, or tunnel_id must be provided.
pattern^tunnel_[a-z0-9]{32}$ tracing: optional "auto" or TracingConfiguration or null Realtime API can write session traces to the Traces Dashboard. Set to null to disable tracing. Once
tracing is enabled for a session, the configuration cannot be modified.
auto will create a trace for the session with default values for the
workflow name, group id, and metadata.
One of the following: Auto = "auto" Enables tracing and sets default values for tracing configuration options. Always auto.
TracingConfiguration { group_id, metadata, workflow_name } Granular configuration for tracing.
group_id: optional string The group id to attach to this trace to enable filtering and
grouping in the Traces Dashboard.
metadata: optional unknown The arbitrary metadata to attach to this trace to enable
filtering in the Traces Dashboard.
workflow_name: optional string The name of the workflow to attach to this trace. This is used to
name the trace in the Traces Dashboard.
truncation: optional RealtimeTruncation When the number of tokens in a conversation exceeds the model’s input token limit, the conversation be truncated, meaning messages (starting from the oldest) will not be included in the model’s context. A 32k context model with 4,096 max output tokens can only include 28,224 tokens in the context before truncation occurs.
Clients can configure truncation behavior to truncate with a lower max token limit, which is an effective way to control token usage and cost.
Truncation will reduce the number of cached tokens on the next turn (busting the cache), since messages are dropped from the beginning of the context. However, clients can also configure truncation to retain messages up to a fraction of the maximum context size, which will reduce the need for future truncations and thus improve the cache rate.
Truncation can be disabled entirely, which means the server will never truncate but would instead return an error if the conversation exceeds the model’s input token limit.
One of the following: "auto" or "disabled" The truncation strategy to use for the session. auto is the default truncation strategy. disabled will disable truncation and emit errors when the conversation exceeds the input token limit.
One of the following: "auto" "disabled" RetentionRatioTruncation { retention_ratio, type, token_limits } Retain a fraction of the conversation tokens when the conversation exceeds the input token limit. This allows you to amortize truncations across multiple turns, which can help improve cached token usage.
retention_ratio: number Fraction of post-instruction conversation tokens to retain (0.0 - 1.0) when the conversation exceeds the input token limit. Setting this to 0.8 means that messages will be dropped until 80% of the maximum allowed tokens are used. This helps reduce the frequency of truncations and improve cache rates.
minimum0 maximum1 type: "retention_ratio" Use retention ratio truncation.
token_limits: optional { post_instructions } Optional custom token limits for this truncation strategy. If not provided, the model’s default token limits will be used.
post_instructions: optional integer Maximum tokens allowed in the conversation after instructions (which including tool definitions). For example, setting this to 5,000 would mean that truncation would occur when the conversation exceeds 5,000 tokens after instructions. This cannot be higher than the model’s context window size minus the maximum output tokens.
minimum0 RealtimeTranscriptionSessionCreateResponse { id, object, type, 3 more } A Realtime transcription session configuration object.
id: string Unique identifier for the session that looks like sess_1234567890abcdef.
object: string The object type. Always realtime.transcription_session.
type: "transcription" The type of session. Always transcription for transcription sessions.
audio: optional { input } Configuration for input audio for the session.
input: optional { format, noise_reduction, transcription, turn_detection } format: optional RealtimeAudioFormats The PCM audio format. Only a 24kHz sample rate is supported.
One of the following: PCMAudio { rate, type } The PCM audio format. Only a 24kHz sample rate is supported.
rate: optional 24000 The sample rate of the audio. Always 24000.
type: optional "audio/pcm" The audio format. Always audio/pcm.
PCMUAudio { type } The G.711 μ-law format.
type: optional "audio/pcmu" The audio format. Always audio/pcmu.
PCMAAudio { type } The G.711 A-law format.
type: optional "audio/pcma" The audio format. Always audio/pcma.
noise_reduction: optional object or null Configuration for input audio noise reduction.
strength: optional number (default: 1) Input noise suppression strength. Higher values apply more suppression.
Defaults to 1 when noise reduction is first enabled. Omitting strength
from a session update preserves the current strength.
minimum0 maximum1 type: optional NoiseReductionType Type of noise reduction. near_field is for close-talking microphones such as headphones, far_field is for far-field microphones such as laptop or conference room microphones.
One of the following: "near_field" "far_field" transcription: optional object or null Configuration of the transcription model.
keywords: optional array of string Words or phrases configured to guide transcription of the input audio.
language: optional string or null The language of the input audio.
languages: optional array of string The possible input audio languages configured for transcription, in ISO-639-1 format.
minItems1 model: optional string or "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model used for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper.
One of the following: string "whisper-1" or "gpt-transcribe" or "gpt-live-transcribe" or 5 more The model used for transcription. Current options are whisper-1, gpt-transcribe, gpt-live-transcribe, gpt-4o-mini-transcribe, gpt-4o-mini-transcribe-2025-12-15, gpt-4o-transcribe, gpt-4o-transcribe-diarize, and gpt-realtime-whisper.
One of the following: "gpt-transcribe" "gpt-live-transcribe" "gpt-realtime-whisper" "whisper-1" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-mini-transcribe-2025-12-15" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
"gpt-4o-transcribe-diarize" Deprecated. Announced shutdown date: 2027-02-26. See https://developers.openai.com/api/docs/deprecations for details and recommended replacements.
prompt: optional string The prompt configured for input audio transcription, when present.
turn_detection: optional RealtimeTranscriptionSessionTurnDetection or null Configuration for turn detection. Can be set to null to turn off. Server
VAD means that the model will detect the start and end of speech based on
audio volume and respond at the end of user speech. For gpt-realtime-whisper, this must be null; VAD is not supported.
prefix_padding_ms: optional integer Amount of audio to include before the VAD detected speech (in
milliseconds). Defaults to 300ms.
silence_duration_ms: optional integer Duration of silence to detect speech stop (in milliseconds). Defaults
to 500ms. With shorter values the model will respond more quickly,
but may jump in on short pauses from the user.
threshold: optional number Activation threshold for VAD (0.0 to 1.0), this defaults to 0.5. A
higher threshold will require louder audio to activate the model, and
thus might perform better in noisy environments.
type: optional string Type of turn detection, only server_vad is currently supported.
expires_at: optional integer Expiration timestamp for the session, in seconds since epoch.
formatunixtime include: optional array of "item.input_audio_transcription.logprobs" or null Additional fields to include in server outputs.
item.input_audio_transcription.logprobs: Include logprobs for input audio transcription.
Returns Examples
Response example
{
"value": "ek_68af296e8e408191a1120ab6383263c2",
"expires_at": 1756310470,
"session": {
"type": "realtime",
"object": "realtime.session",
"id": "sess_C9CiUVUzUzYIssh3ELY1d",
"model": "gpt-realtime",
"output_modalities": [
"audio"
],
"instructions": "You are a friendly assistant.",
"tools": [],
"tool_choice": "auto",
"max_output_tokens": "inf",
"tracing": null,
"truncation": "auto",
"prompt": null,
"expires_at": 0,
"audio": {
"input": {
"format": {
"type": "audio/pcm",
"rate": 24000
},
"transcription": null,
"noise_reduction": null,
"turn_detection": {
"type": "server_vad"
}
},
"output": {
"format": {
"type": "audio/pcm",
"rate": 24000
},
"voice": "alloy",
"speed": 1.0
}
},
"include": null
}
}