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

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:



504
505
506
507
 
# File 'lib/openai/resources/chat/completions.rb', line 504

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:



12
13
14
 
# File 'lib/openai/resources/chat/completions.rb', line 12

def messages
  @messages
end

Instance Method Details

#build_tools_with_models(tools, tool_models) ⇒ Object 



228
229
230
231
232
233
234
235
236
237
238
239
 
# File 'lib/openai/resources/chat/completions.rb', line 228

def build_tools_with_models(tools, tool_models)
  return [] if tools.nil?

  tools.map do |tool|
    next tool unless tool[:type] == :function

    function_name = tool.dig(:function, :name)
    model = tool_models[function_name]

    model ? tool.merge(model: model) : tool
  end
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_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 #stream_raw for streaming counterpart.

Some parameter documentations has been truncated, see Models::Chat::CompletionCreateParams for more details.

**Starting a new project?** We recommend trying [Responses](platform.openai.com/docs/api-reference/responses) to take advantage of the latest OpenAI platform features. Compare [Chat Completions with Responses](platform.openai.com/docs/guides/responses-vs-chat-completions?api-mode=responses).

Creates a model response for the given chat conversation. Learn more in the [text generation](platform.openai.com/docs/guides/text-generation), [vision](platform.openai.com/docs/guides/vision), and [audio](platform.openai.com/docs/guides/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](platform.openai.com/docs/guides/reasoning).

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

Parameters:

Returns:

See Also:



117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
 
# File 'lib/openai/resources/chat/completions.rb', line 117

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)

  # rubocop:disable Metrics/BlockLength
  unwrap = ->(raw) do
    if model.is_a?(OpenAI::StructuredOutput::JsonSchemaConverter)
      raw[:choices]&.each do |choice|
        message = choice.fetch(:message)
        begin
          content = message.fetch(:content)
          parsed = content.nil? ? nil : JSON.parse(content, symbolize_names: true)
        rescue JSON::ParserError => e
          parsed = e
        end
        coerced = OpenAI::Internal::Type::Converter.coerce(model, parsed)
        message.store(:parsed, coerced)
      end
    end
    raw[:choices]&.each do |choice|
      choice.dig(:message, :tool_calls)&.each do |tool_call|
        func = tool_call.fetch(:function)
        next if (model = tool_models[func.fetch(:name)]).nil?

        begin
          arguments = func.fetch(:arguments)
          parsed = arguments.nil? ? nil : JSON.parse(arguments, symbolize_names: true)
        rescue JSON::ParserError => e
          parsed = e
        end
        coerced = OpenAI::Internal::Type::Converter.coerce(model, parsed)
        func.store(:parsed, coerced)
      end
    end

    raw
  end
  # rubocop:enable Metrics/BlockLength

  @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:



491
492
493
494
495
496
497
498
499
 
# File 'lib/openai/resources/chat/completions.rb', line 491

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 



172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
 
# File 'lib/openai/resources/chat/completions.rb', line 172

def get_structured_output_models(parsed)
  model = nil
  tool_models = {}
  case parsed
  in {response_format: OpenAI::StructuredOutput::JsonSchemaConverter => model}
    parsed.update(
      response_format: {
        type: :json_schema,
        json_schema: {
          strict: true,
          name: model.name.split("::").last,
          schema: model.to_json_schema
        }
      }
    )
  in {response_format: {type: :json_schema, json_schema: OpenAI::StructuredOutput::JsonSchemaConverter => model}}
    parsed.fetch(:response_format).update(
      json_schema: {
        strict: true,
        name: model.name.split("::").last,
        schema: model.to_json_schema
      }
    )
  in {response_format: {type: :json_schema, json_schema: {schema: OpenAI::StructuredOutput::JsonSchemaConverter => model}}}
    parsed.dig(:response_format, :json_schema).store(:schema, model.to_json_schema)
  in {tools: Array => tools}
    mapped = tools.map do |tool|
      case tool
      in OpenAI::StructuredOutput::JsonSchemaConverter
        name = tool.name.split("::").last
        tool_models.store(name, tool)
        {
          type: :function,
          function: {
            strict: true,
            name: name,
            parameters: tool.to_json_schema
          }
        }
      in {function: {parameters: OpenAI::StructuredOutput::JsonSchemaConverter => params}}
        func = tool.fetch(:function)
        name = func[:name] ||= params.name.split("::").last
        tool_models.store(name, params)
        func.update(parameters: params.to_json_schema)
        tool
      else
        tool
      end
    end
    tools.replace(mapped)
  else
  end

  [model, tool_models]
end

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

Some parameter documentations has been truncated, see Models::Chat::CompletionListParams for more details.

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:

  • 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 `

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

Returns:

See Also:



465
466
467
468
469
470
471
472
473
474
475
476
477
 
# File 'lib/openai/resources/chat/completions.rb', line 465

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:



402
403
404
405
406
407
408
409
410
 
# File 'lib/openai/resources/chat/completions.rb', line 402

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 



241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
 
# File 'lib/openai/resources/chat/completions.rb', line 241

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_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 #create for non-streaming counterpart.

Some parameter documentations has been truncated, see Models::Chat::CompletionCreateParams for more details.

**Starting a new project?** We recommend trying [Responses](platform.openai.com/docs/api-reference/responses) to take advantage of the latest OpenAI platform features. Compare [Chat Completions with Responses](platform.openai.com/docs/guides/responses-vs-chat-completions?api-mode=responses).

Creates a model response for the given chat conversation. Learn more in the [text generation](platform.openai.com/docs/guides/text-generation), [vision](platform.openai.com/docs/guides/vision), and [audio](platform.openai.com/docs/guides/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](platform.openai.com/docs/guides/reasoning).

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

Parameters:

Returns:

See Also:



371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
 
# File 'lib/openai/resources/chat/completions.rb', line 371

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

Some parameter documentations has been truncated, see Models::Chat::CompletionUpdateParams for more details.

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

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

Returns:

See Also:



430
431
432
433
434
435
436
437
438
439
440
 
# File 'lib/openai/resources/chat/completions.rb', line 430

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