Class: OpenAI::Resources::Chat::Completions

Inherits:
Object
  • Object
show all
Defined in:
lib/openai/resources/chat/completions.rb,
lib/openai/resources/chat/completions/messages.rb,
sig/openai/helpers/sorbet.rbs,
sig/openai/resources/chat/completions.rbs,
sig/openai/resources/chat/completions/messages.rbs

Overview

Given a list of messages comprising a conversation, the model will return a response.

Defined Under Namespace

Classes: Messages

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(client:) ⇒ Completions

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Returns a new instance of Completions.

Parameters:



842
843
844
845
 
# File 'lib/openai/resources/chat/completions.rb', line 842

def initialize(client:)
  @client = client
  @messages = OpenAI::Resources::Chat::Completions::Messages.new(client: client)
end

Instance Attribute Details

#messages ⇒ OpenAI::Resources::Chat::Completions::Messages (readonly)

Given a list of messages comprising a conversation, the model will return a response.

Returns:



14
15
16
 
# File 'lib/openai/resources/chat/completions.rb', line 14

def messages
  @messages
end

Instance Method Details

#build_tools_with_models(tools, tool_models) ⇒ Object 



354
355
356
 
# File 'lib/openai/resources/chat/completions.rb', line 354

def build_tools_with_models(tools, tool_models)
  OpenAI::Helpers::StructuredOutput::ChatCompletionParser.build_tools(tools, tool_models)
end

#create(messages:, model:, audio: nil, frequency_penalty: nil, function_call: nil, functions: nil, logit_bias: nil, logprobs: nil, max_completion_tokens: nil, max_tokens: nil, metadata: nil, modalities: nil, moderation: nil, n: nil, parallel_tool_calls: nil, prediction: nil, presence_penalty: nil, prompt_cache_key: nil, prompt_cache_options: nil, prompt_cache_retention: nil, reasoning_effort: nil, response_format: nil, safety_identifier: nil, seed: nil, service_tier: nil, stop: nil, store: nil, stream_options: nil, temperature: nil, tool_choice: nil, tools: nil, top_logprobs: nil, top_p: nil, user: nil, verbosity: nil, web_search_options: nil, request_options: {}) ⇒ OpenAI::Models::Chat::ChatCompletion

See OpenAI::Resources::Chat::Completions#stream_raw for streaming counterpart.

Starting a new project? We recommend trying Responses to take advantage of the latest OpenAI platform features. Compare Chat Completions with Responses.

Creates a model response for the given chat conversation. Learn more in the text generation, vision, and audio guides.

Parameter support can differ depending on the model used to generate the response, particularly for newer reasoning models. Parameters that are only supported for reasoning models are noted below. For the current state of unsupported parameters in reasoning models, refer to the reasoning guide.

Returns a chat completion object, or a streamed sequence of chat completion chunk objects if the request is streamed.

Parameters:

  • messages (Array<OpenAI::Models::Chat::ChatCompletionDeveloperMessageParam, OpenAI::Models::Chat::ChatCompletionSystemMessageParam, OpenAI::Models::Chat::ChatCompletionUserMessageParam, OpenAI::Models::Chat::ChatCompletionAssistantMessageParam, OpenAI::Models::Chat::ChatCompletionToolMessageParam, OpenAI::Models::Chat::ChatCompletionFunctionMessageParam>) —

    A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, images, and audio.

  • model (String, Symbol, OpenAI::Models::ChatModel) —

    Model ID used to generate the response, like gpt-6-astra or o3. OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the model guide to browse and compare available models.

  • audio (OpenAI::Models::Chat::ChatCompletionAudioParam, nil) —

    Parameters for audio output. Required when audio output is requested with modalities: ["audio"]. Learn more.

  • frequency_penalty (Float, nil) —

    Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

  • function_call (Symbol, OpenAI::Models::Chat::CompletionCreateParams::FunctionCall::FunctionCallMode, OpenAI::Models::Chat::ChatCompletionFunctionCallOption) —

    Deprecated in favor of tool_choice.

    Controls which (if any) function is called by the model.

    none means the model will not call a function and instead generates a message.

    auto means the model can pick between generating a message or calling a function.

    Specifying a particular function via {"name": "my_function"} forces the model to call that function.

    none is the default when no functions are present. auto is the default if functions are present.

  • functions (Array<OpenAI::Models::Chat::CompletionCreateParams::Function>) —

    Deprecated in favor of tools.

    A list of functions the model may generate JSON inputs for.

  • logit_bias (Hash{Symbol=>Integer}, nil) —

    Modify the likelihood of specified tokens appearing in the completion.

    Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.

  • logprobs (Boolean, nil) —

    Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.

  • max_completion_tokens (Integer, nil) —

    An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

  • max_tokens (Integer, nil) —

    The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

    This value is now deprecated in favor of max_completion_tokens, and is not compatible with o-series models.

  • metadata (Hash{Symbol=>String}, nil) —

    Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • modalities (Array<Symbol, OpenAI::Models::Chat::CompletionCreateParams::Modality>, nil) —

    Output types that you would like the model to generate. Most models are capable of generating text, which is the default:

    ["text"]

    The gpt-4o-audio-preview model can also be used to generate audio. To request that this model generate both text and audio responses, you can use:

    ["text", "audio"]

  • moderation (OpenAI::Models::Chat::CompletionCreateParams::Moderation, nil) —

    Configuration for running moderation on the request input and generated output.

  • n (Integer, nil) —

    How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

  • parallel_tool_calls (Boolean) —

    Whether to enable parallel function calling during tool use.

  • prediction (OpenAI::Models::Chat::ChatCompletionPredictionContent, nil) —

    Static predicted output content, such as the content of a text file that is being regenerated.

  • presence_penalty (Float, nil) —

    Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

  • prompt_cache_key (String, nil) —

    Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the user field. Learn more.

  • prompt_cache_options (OpenAI::Models::Chat::CompletionCreateParams::PromptCacheOptions) —

    Options for prompt caching. Supported for gpt-5.6 and later models. By default, OpenAI automatically chooses one implicit cache breakpoint. You can add explicit breakpoints to content blocks with prompt_cache_breakpoint. Each request can write up to four breakpoints. For cache matching, OpenAI considers up to the latest 80 breakpoints in the conversation, without a content-block lookback limit. Set mode to explicit to disable the implicit breakpoint. The ttl defaults to 30m, which is currently the only supported value. See the prompt caching guide for current details.

  • prompt_cache_retention (Symbol, OpenAI::Models::Chat::CompletionCreateParams::PromptCacheRetention, nil) —

    Deprecated. Use prompt_cache_options.ttl instead.

    The retention policy for the prompt cache. Set to 24h to enable extended prompt caching, which keeps cached prefixes active for longer, up to a maximum of 24 hours. Learn more. This field expresses a maximum retention policy, while prompt_cache_options.ttl expresses a minimum cache lifetime. The two fields are independent and do not interact. For gpt-5.5, gpt-5.5-pro, and future models, only 24h is supported.

    For older models that support both in_memory and 24h, the default depends on your organization's data retention policy:

    • Organizations without ZDR enabled default to 24h.
    • Organizations with ZDR enabled default to in_memory when prompt_cache_retention is not specified.
  • reasoning_effort (Symbol, OpenAI::Models::ReasoningEffort, nil) —

    Constrains effort on reasoning for reasoning models. Currently supported values are none, minimal, low, medium, high, xhigh, and max. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response. Not all reasoning models support every value. See the reasoning guide for model-specific support.

  • response_format (OpenAI::Models::ResponseFormatText, OpenAI::Models::ResponseFormatJSONSchema, OpenAI::Models::ResponseFormatJSONObject) —

    An object specifying the format that the model must output.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

  • safety_identifier (String, nil) —

    A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. Learn more.

  • seed (Integer, nil) —

    This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result. Determinism is not guaranteed, and you should refer to the system_fingerprint response parameter to monitor changes in the backend.

  • service_tier (Symbol, OpenAI::Models::Chat::CompletionCreateParams::ServiceTier, nil) —

    Specifies the processing type used for serving the request.

    • If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'.
    • If set to 'default', then the request will be processed with the standard pricing and performance for the selected model.
    • If set to 'flex', then the request will be processed with the Flex Processing service tier.
    • To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request.
    • When not set, the default behavior is 'auto'.

    When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

  • stop (String, Array<String>, nil) —

    Not supported with latest reasoning models o3 and o4-mini.

    Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

  • store (Boolean, nil) —

    Whether or not to store the output of this chat completion request for use in our model distillation or evals products.

    Supports text and image inputs. Note: image inputs over 8MB will be dropped.

  • stream_options (OpenAI::Models::Chat::ChatCompletionStreamOptions, nil) —

    Options for streaming response. Only set this when you set stream: true.

  • temperature (Float, nil) —

    What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or top_p but not both.

  • tool_choice (Symbol, OpenAI::Models::Chat::ChatCompletionToolChoiceOption::Auto, OpenAI::Models::Chat::ChatCompletionAllowedToolChoice, OpenAI::Models::Chat::ChatCompletionNamedToolChoice, OpenAI::Models::Chat::ChatCompletionNamedToolChoiceCustom) —

    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. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool.

    none is the default when no tools are present. auto is the default if tools are present.

  • tools (Array<OpenAI::Models::Chat::ChatCompletionFunctionTool, OpenAI::Models::Chat::ChatCompletionCustomTool>) —

    A list of tools the model may call. You can provide either custom tools or function tools.

  • top_logprobs (Integer, nil) —

    An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. logprobs must be set to true if this parameter is used.

  • top_p (Float, nil) —

    An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

  • user (String) —

    This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

  • verbosity (Symbol, OpenAI::Models::Chat::CompletionCreateParams::Verbosity, nil) —

    Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

  • web_search_options (OpenAI::Models::Chat::CompletionCreateParams::WebSearchOptions) —

    This tool searches the web for relevant results to use in a response. Learn more about the web search tool.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
 
# File 'lib/openai/resources/chat/completions.rb', line 328

def create(params)
  parsed, options = OpenAI::Chat::CompletionCreateParams.dump_request(params)
  if parsed[:stream]
    message = "Please use `#stream_raw` for the streaming use case."
    raise ArgumentError.new(message)
  end

  model, tool_models = get_structured_output_models(parsed)

  unwrap = OpenAI::Helpers::StructuredOutput::ChatCompletionParser.build_unwrap(model, tool_models)

  @client.request(
    method: :post,
    path: "chat/completions",
    body: parsed,
    unwrap: unwrap,
    model: OpenAI::Chat::ChatCompletion,
    security: {bearer_auth: true},
    options: options
  )
end

#delete(completion_id, request_options: {}) ⇒ OpenAI::Models::Chat::ChatCompletionDeleted

Delete a stored chat completion. Only Chat Completions that have been created with the store parameter set to true can be deleted.

Parameters:

  • completion_id (String) —

    The ID of the chat completion to delete.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



829
830
831
832
833
834
835
836
837
 
# File 'lib/openai/resources/chat/completions.rb', line 829

def delete(completion_id, params = {})
  @client.request(
    method: :delete,
    path: ["chat/completions/%1$s", completion_id],
    model: OpenAI::Chat::ChatCompletionDeleted,
    security: {bearer_auth: true},
    options: params[:request_options]
  )
end

#get_structured_output_models(parsed) ⇒ Object 



350
351
352
 
# File 'lib/openai/resources/chat/completions.rb', line 350

def get_structured_output_models(parsed)
  OpenAI::Helpers::StructuredOutput::ChatCompletionParser.get_models(parsed)
end

#list(after: nil, limit: nil, metadata: nil, model: nil, order: nil, request_options: {}) ⇒ OpenAI::Internal::CursorPage<OpenAI::Models::Chat::ChatCompletion>

List stored Chat Completions. Only Chat Completions that have been stored with the store parameter set to true will be returned.

Parameters:

  • after (String) —

    Identifier for the last chat completion from the previous pagination request.

  • limit (Integer) —

    Number of Chat Completions to retrieve.

  • metadata (Hash{Symbol=>String}, nil) —

    A list of metadata keys to filter the Chat Completions by. Example:

    metadata[key1]=value1&metadata[key2]=value2

  • model (String) —

    The model used to generate the Chat Completions.

  • order (Symbol, OpenAI::Models::Chat::CompletionListParams::Order) —

    Sort order for Chat Completions by timestamp. Use asc for ascending order or desc for descending order. Defaults to asc.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



802
803
804
805
806
807
808
809
810
811
812
813
814
 
# File 'lib/openai/resources/chat/completions.rb', line 802

def list(params = {})
  parsed, options = OpenAI::Chat::CompletionListParams.dump_request(params)
  query = OpenAI::Internal::Util.encode_query_params(parsed)
  @client.request(
    method: :get,
    path: "chat/completions",
    query: query,
    page: OpenAI::Internal::CursorPage,
    model: OpenAI::Chat::ChatCompletion,
    security: {bearer_auth: true},
    options: options
  )
end

#retrieve(completion_id, request_options: {}) ⇒ OpenAI::Models::Chat::ChatCompletion

Get a stored chat completion. Only Chat Completions that have been created with the store parameter set to true will be returned.

Parameters:

  • completion_id (String) —

    The ID of the chat completion to retrieve.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



730
731
732
733
734
735
736
737
738
 
# File 'lib/openai/resources/chat/completions.rb', line 730

def retrieve(completion_id, params = {})
  @client.request(
    method: :get,
    path: ["chat/completions/%1$s", completion_id],
    model: OpenAI::Chat::ChatCompletion,
    security: {bearer_auth: true},
    options: params[:request_options]
  )
end

#stream(params) ⇒ Object 



358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
 
# File 'lib/openai/resources/chat/completions.rb', line 358

def stream(params)
  parsed, options = OpenAI::Chat::CompletionCreateParams.dump_request(params)

  parsed.store(:stream, true)

  response_format, tool_models = get_structured_output_models(parsed)

  input_tools = build_tools_with_models(parsed[:tools], tool_models)

  raw_stream = @client.request(
    method: :post,
    path: "chat/completions",
    headers: {"accept" => "text/event-stream"},
    body: parsed,
    stream: OpenAI::Internal::Stream,
    model: OpenAI::Chat::ChatCompletionChunk,
    security: {bearer_auth: true},
    options: options
  )

  OpenAI::Helpers::Streaming::ChatCompletionStream.new(
    raw_stream: raw_stream,
    response_format: response_format,
    input_tools: input_tools
  )
end

#stream_raw(messages:, model:, audio: nil, frequency_penalty: nil, function_call: nil, functions: nil, logit_bias: nil, logprobs: nil, max_completion_tokens: nil, max_tokens: nil, metadata: nil, modalities: nil, moderation: nil, n: nil, parallel_tool_calls: nil, prediction: nil, presence_penalty: nil, prompt_cache_key: nil, prompt_cache_options: nil, prompt_cache_retention: nil, reasoning_effort: nil, response_format: nil, safety_identifier: nil, seed: nil, service_tier: nil, stop: nil, store: nil, stream_options: nil, temperature: nil, tool_choice: nil, tools: nil, top_logprobs: nil, top_p: nil, user: nil, verbosity: nil, web_search_options: nil, request_options: {}) ⇒ OpenAI::Internal::Stream<OpenAI::Models::Chat::ChatCompletionChunk>

See OpenAI::Resources::Chat::Completions#create for non-streaming counterpart.

Starting a new project? We recommend trying Responses to take advantage of the latest OpenAI platform features. Compare Chat Completions with Responses.

Creates a model response for the given chat conversation. Learn more in the text generation, vision, and audio guides.

Parameter support can differ depending on the model used to generate the response, particularly for newer reasoning models. Parameters that are only supported for reasoning models are noted below. For the current state of unsupported parameters in reasoning models, refer to the reasoning guide.

Returns a chat completion object, or a streamed sequence of chat completion chunk objects if the request is streamed.

Parameters:

  • messages (Array<OpenAI::Models::Chat::ChatCompletionDeveloperMessageParam, OpenAI::Models::Chat::ChatCompletionSystemMessageParam, OpenAI::Models::Chat::ChatCompletionUserMessageParam, OpenAI::Models::Chat::ChatCompletionAssistantMessageParam, OpenAI::Models::Chat::ChatCompletionToolMessageParam, OpenAI::Models::Chat::ChatCompletionFunctionMessageParam>) —

    A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, images, and audio.

  • model (String, Symbol, OpenAI::Models::ChatModel) —

    Model ID used to generate the response, like gpt-6-astra or o3. OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the model guide to browse and compare available models.

  • audio (OpenAI::Models::Chat::ChatCompletionAudioParam, nil) —

    Parameters for audio output. Required when audio output is requested with modalities: ["audio"]. Learn more.

  • frequency_penalty (Float, nil) —

    Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

  • function_call (Symbol, OpenAI::Models::Chat::CompletionCreateParams::FunctionCall::FunctionCallMode, OpenAI::Models::Chat::ChatCompletionFunctionCallOption) —

    Deprecated in favor of tool_choice.

    Controls which (if any) function is called by the model.

    none means the model will not call a function and instead generates a message.

    auto means the model can pick between generating a message or calling a function.

    Specifying a particular function via {"name": "my_function"} forces the model to call that function.

    none is the default when no functions are present. auto is the default if functions are present.

  • functions (Array<OpenAI::Models::Chat::CompletionCreateParams::Function>) —

    Deprecated in favor of tools.

    A list of functions the model may generate JSON inputs for.

  • logit_bias (Hash{Symbol=>Integer}, nil) —

    Modify the likelihood of specified tokens appearing in the completion.

    Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.

  • logprobs (Boolean, nil) —

    Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.

  • max_completion_tokens (Integer, nil) —

    An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

  • max_tokens (Integer, nil) —

    The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

    This value is now deprecated in favor of max_completion_tokens, and is not compatible with o-series models.

  • metadata (Hash{Symbol=>String}, nil) —

    Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • modalities (Array<Symbol, OpenAI::Models::Chat::CompletionCreateParams::Modality>, nil) —

    Output types that you would like the model to generate. Most models are capable of generating text, which is the default:

    ["text"]

    The gpt-4o-audio-preview model can also be used to generate audio. To request that this model generate both text and audio responses, you can use:

    ["text", "audio"]

  • moderation (OpenAI::Models::Chat::CompletionCreateParams::Moderation, nil) —

    Configuration for running moderation on the request input and generated output.

  • n (Integer, nil) —

    How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

  • parallel_tool_calls (Boolean) —

    Whether to enable parallel function calling during tool use.

  • prediction (OpenAI::Models::Chat::ChatCompletionPredictionContent, nil) —

    Static predicted output content, such as the content of a text file that is being regenerated.

  • presence_penalty (Float, nil) —

    Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

  • prompt_cache_key (String, nil) —

    Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the user field. Learn more.

  • prompt_cache_options (OpenAI::Models::Chat::CompletionCreateParams::PromptCacheOptions) —

    Options for prompt caching. Supported for gpt-5.6 and later models. By default, OpenAI automatically chooses one implicit cache breakpoint. You can add explicit breakpoints to content blocks with prompt_cache_breakpoint. Each request can write up to four breakpoints. For cache matching, OpenAI considers up to the latest 80 breakpoints in the conversation, without a content-block lookback limit. Set mode to explicit to disable the implicit breakpoint. The ttl defaults to 30m, which is currently the only supported value. See the prompt caching guide for current details.

  • prompt_cache_retention (Symbol, OpenAI::Models::Chat::CompletionCreateParams::PromptCacheRetention, nil) —

    Deprecated. Use prompt_cache_options.ttl instead.

    The retention policy for the prompt cache. Set to 24h to enable extended prompt caching, which keeps cached prefixes active for longer, up to a maximum of 24 hours. Learn more. This field expresses a maximum retention policy, while prompt_cache_options.ttl expresses a minimum cache lifetime. The two fields are independent and do not interact. For gpt-5.5, gpt-5.5-pro, and future models, only 24h is supported.

    For older models that support both in_memory and 24h, the default depends on your organization's data retention policy:

    • Organizations without ZDR enabled default to 24h.
    • Organizations with ZDR enabled default to in_memory when prompt_cache_retention is not specified.
  • reasoning_effort (Symbol, OpenAI::Models::ReasoningEffort, nil) —

    Constrains effort on reasoning for reasoning models. Currently supported values are none, minimal, low, medium, high, xhigh, and max. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response. Not all reasoning models support every value. See the reasoning guide for model-specific support.

  • response_format (OpenAI::Models::ResponseFormatText, OpenAI::Models::ResponseFormatJSONSchema, OpenAI::Models::ResponseFormatJSONObject) —

    An object specifying the format that the model must output.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

  • safety_identifier (String, nil) —

    A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. Learn more.

  • seed (Integer, nil) —

    This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result. Determinism is not guaranteed, and you should refer to the system_fingerprint response parameter to monitor changes in the backend.

  • service_tier (Symbol, OpenAI::Models::Chat::CompletionCreateParams::ServiceTier, nil) —

    Specifies the processing type used for serving the request.

    • If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'.
    • If set to 'default', then the request will be processed with the standard pricing and performance for the selected model.
    • If set to 'flex', then the request will be processed with the Flex Processing service tier.
    • To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request.
    • When not set, the default behavior is 'auto'.

    When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

  • stop (String, Array<String>, nil) —

    Not supported with latest reasoning models o3 and o4-mini.

    Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

  • store (Boolean, nil) —

    Whether or not to store the output of this chat completion request for use in our model distillation or evals products.

    Supports text and image inputs. Note: image inputs over 8MB will be dropped.

  • stream_options (OpenAI::Models::Chat::ChatCompletionStreamOptions, nil) —

    Options for streaming response. Only set this when you set stream: true.

  • temperature (Float, nil) —

    What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or top_p but not both.

  • tool_choice (Symbol, OpenAI::Models::Chat::ChatCompletionToolChoiceOption::Auto, OpenAI::Models::Chat::ChatCompletionAllowedToolChoice, OpenAI::Models::Chat::ChatCompletionNamedToolChoice, OpenAI::Models::Chat::ChatCompletionNamedToolChoiceCustom) —

    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. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool.

    none is the default when no tools are present. auto is the default if tools are present.

  • tools (Array<OpenAI::Models::Chat::ChatCompletionFunctionTool, OpenAI::Models::Chat::ChatCompletionCustomTool>) —

    A list of tools the model may call. You can provide either custom tools or function tools.

  • top_logprobs (Integer, nil) —

    An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. logprobs must be set to true if this parameter is used.

  • top_p (Float, nil) —

    An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

  • user (String) —

    This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

  • verbosity (Symbol, OpenAI::Models::Chat::CompletionCreateParams::Verbosity, nil) —

    Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

  • web_search_options (OpenAI::Models::Chat::CompletionCreateParams::WebSearchOptions) —

    This tool searches the web for relevant results to use in a response. Learn more about the web search tool.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
 
# File 'lib/openai/resources/chat/completions.rb', line 697

def stream_raw(params)
  parsed, options = OpenAI::Chat::CompletionCreateParams.dump_request(params)
  unless parsed.fetch(:stream, true)
    message = "Please use `#create` for the non-streaming use case."
    raise ArgumentError.new(message)
  end

  parsed.store(:stream, true)
  @client.request(
    method: :post,
    path: "chat/completions",
    headers: {"accept" => "text/event-stream", "accept-encoding" => "identity"},
    body: parsed,
    stream: OpenAI::Internal::Stream,
    model: OpenAI::Chat::ChatCompletionChunk,
    security: {bearer_auth: true},
    options: options
  )
end

#update(completion_id, metadata:, request_options: {}) ⇒ OpenAI::Models::Chat::ChatCompletion

Modify a stored chat completion. Only Chat Completions that have been created with the store parameter set to true can be modified. Currently, the only supported modification is to update the metadata field.

Parameters:

  • completion_id (String) —

    The ID of the chat completion to update.

  • metadata (Hash{Symbol=>String}, nil) —

    Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • request_options (OpenAI::RequestOptions, Hash{Symbol=>Object}, nil)

Returns:

See Also:



762
763
764
765
766
767
768
769
770
771
772
 
# File 'lib/openai/resources/chat/completions.rb', line 762

def update(completion_id, params)
  parsed, options = OpenAI::Chat::CompletionUpdateParams.dump_request(params)
  @client.request(
    method: :post,
    path: ["chat/completions/%1$s", completion_id],
    body: parsed,
    model: OpenAI::Chat::ChatCompletion,
    security: {bearer_auth: true},
    options: options
  )
end