openai 0.31.0 → 0.32.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2c2acb13f76000aa282621830c5724ffe7ceebf464e7cd4d58b4d8b99d6ddd5d
-  data.tar.gz: e55e41d0e9dd00daa7ab4dd6409ea8f4fc68c91ccf09b8660725634ab563c56b
+  metadata.gz: 93eb98c219356a4e424930695d90aad0d6100f2d7547416e1ab30ccc7527e69a
+  data.tar.gz: 6fe4f2e35ec1f2ce294eb14d92fd8fff1e872d1e2f2c93e33b0f0a7014bdda6b
 SHA512:
-  metadata.gz: ffb82dfdebeca385d0f584be09ea4bbf5ec79e0031209f362a93fc1167f78b04ddc2743e77ad18f87f9e9b1a33909f9d532fb5229379139479ef09317b263289
-  data.tar.gz: dd09d9967b4cb16fd70bbbcc153ab34bf144a1dd3e922100f3f92fa597a3756fcb112763392745934b647d61ba48925070e55889a09a0b94ee89ff435f326cc7
+  metadata.gz: e13c1e23f3e17b8a8cbf52696205fbe81ab4d2d369d61bb327acaf9dad77673ab8adf636b109a117bd8ccf9fb619081a5e63d37493f85cfe5839a52a78653024
+  data.tar.gz: 1ace0e832f9f9542b5d8bdd968d8cf127b9501d074846e66edcd265e362fbcef5a83a7a59d5feb4a064ab4e76c1271109070cbf2da9e3dc2d249d79691ca395e

data/CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## 0.32.0 (2025-10-16)
+
+Full Changelog: [v0.31.0...v0.32.0](https://github.com/openai/openai-ruby/compare/v0.31.0...v0.32.0)
+
+### Features
+
+* **api:** Add support for gpt-4o-transcribe-diarize on audio/transcriptions endpoint ([b31bd7f](https://github.com/openai/openai-ruby/commit/b31bd7f20ca702160873fa26ab39479fd8102f85))
+
+
+### Bug Fixes
+
+* absolutely qualified uris should always override the default ([14fdff8](https://github.com/openai/openai-ruby/commit/14fdff8de533a1002c64c9086016777a1e152a97))
+* should not reuse buffers for `IO.copy_stream` interop ([8f33de1](https://github.com/openai/openai-ruby/commit/8f33de18bb104d5003a4d459ad244c0813e5a07e))
+
 ## 0.31.0 (2025-10-10)
 
 Full Changelog: [v0.30.0...v0.31.0](https://github.com/openai/openai-ruby/compare/v0.30.0...v0.31.0)

data/README.md

@@ -15,7 +15,7 @@ To use this gem, install via Bundler by adding the following to your application
 <!-- x-release-please-start-version -->
 
 ```ruby
-gem "openai", "~> 0.31.0"
+gem "openai", "~> 0.32.0"
 ```
 
 <!-- x-release-please-end -->

data/lib/openai/internal/util.rb

@@ -346,8 +346,9 @@ module OpenAI
           base_path, base_query = lhs.fetch_values(:path, :query)
           slashed = base_path.end_with?("/") ? base_path : "#{base_path}/"
 
-          parsed_path, parsed_query = parse_uri(rhs.fetch(:path)).fetch_values(:path, :query)
-          override = URI::Generic.build(**rhs.slice(:scheme, :host, :port), path: parsed_path)
+          merged = {**parse_uri(rhs.fetch(:path)), **rhs.except(:path, :query)}
+          parsed_path, parsed_query = merged.fetch_values(:path, :query)
+          override = URI::Generic.build(**merged.slice(:scheme, :host, :port), path: parsed_path)
 
           joined = URI.join(URI::Generic.build(lhs.except(:path, :query)), slashed, override)
           query = deep_merge(
@@ -473,10 +474,9 @@ module OpenAI
         # @return [Enumerable<String>]
         def writable_enum(&blk)
           Enumerator.new do |y|
-            buf = String.new
             y.define_singleton_method(:write) do
-              self << buf.replace(_1)
-              buf.bytesize
+              self << _1.dup
+              _1.bytesize
             end
 
             blk.call(y)

data/lib/openai/models/audio/transcription_create_params.rb

@@ -19,8 +19,8 @@ module OpenAI
 
         # @!attribute model
         #   ID of the model to use. The options are `gpt-4o-transcribe`,
-        #   `gpt-4o-mini-transcribe`, and `whisper-1` (which is powered by our open source
-        #   Whisper V2 model).
+        #   `gpt-4o-mini-transcribe`, `whisper-1` (which is powered by our open source
+        #   Whisper V2 model), and `gpt-4o-transcribe-diarize`.
         #
         #   @return [String, Symbol, OpenAI::Models::AudioModel]
         required :model, union: -> { OpenAI::Audio::TranscriptionCreateParams::Model }
@@ -30,6 +30,8 @@ module OpenAI
         #   first normalizes loudness and then uses voice activity detection (VAD) to choose
         #   boundaries. `server_vad` object can be provided to tweak VAD detection
         #   parameters manually. If unset, the audio is transcribed as a single block.
+        #   Required when using `gpt-4o-transcribe-diarize` for inputs longer than 30
+        #   seconds.
         #
         #   @return [Symbol, :auto, OpenAI::Models::Audio::TranscriptionCreateParams::ChunkingStrategy::VadConfig, nil]
         optional :chunking_strategy,
@@ -41,11 +43,30 @@ module OpenAI
         #   return the log probabilities of the tokens in the response to understand the
         #   model's confidence in the transcription. `logprobs` only works with
         #   response_format set to `json` and only with the models `gpt-4o-transcribe` and
-        #   `gpt-4o-mini-transcribe`.
+        #   `gpt-4o-mini-transcribe`. This field is not supported when using
+        #   `gpt-4o-transcribe-diarize`.
         #
         #   @return [Array<Symbol, OpenAI::Models::Audio::TranscriptionInclude>, nil]
         optional :include, -> { OpenAI::Internal::Type::ArrayOf[enum: OpenAI::Audio::TranscriptionInclude] }
 
+        # @!attribute known_speaker_names
+        #   Optional list of speaker names that correspond to the audio samples provided in
+        #   `known_speaker_references[]`. Each entry should be a short identifier (for
+        #   example `customer` or `agent`). Up to 4 speakers are supported.
+        #
+        #   @return [Array<String>, nil]
+        optional :known_speaker_names, OpenAI::Internal::Type::ArrayOf[String]
+
+        # @!attribute known_speaker_references
+        #   Optional list of audio samples (as
+        #   [data URLs](https://developer.mozilla.org/en-US/docs/Web/HTTP/Basics_of_HTTP/Data_URLs))
+        #   that contain known speaker references matching `known_speaker_names[]`. Each
+        #   sample must be between 2 and 10 seconds, and can use any of the same input audio
+        #   formats supported by `file`.
+        #
+        #   @return [Array<String>, nil]
+        optional :known_speaker_references, OpenAI::Internal::Type::ArrayOf[String]
+
         # @!attribute language
         #   The language of the input audio. Supplying the input language in
         #   [ISO-639-1](https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes) (e.g. `en`)
@@ -58,15 +79,18 @@ module OpenAI
         #   An optional text to guide the model's style or continue a previous audio
         #   segment. The
         #   [prompt](https://platform.openai.com/docs/guides/speech-to-text#prompting)
-        #   should match the audio language.
+        #   should match the audio language. This field is not supported when using
+        #   `gpt-4o-transcribe-diarize`.
         #
         #   @return [String, nil]
         optional :prompt, String
 
         # @!attribute response_format
         #   The format of the output, in one of these options: `json`, `text`, `srt`,
-        #   `verbose_json`, or `vtt`. For `gpt-4o-transcribe` and `gpt-4o-mini-transcribe`,
-        #   the only supported format is `json`.
+        #   `verbose_json`, `vtt`, or `diarized_json`. For `gpt-4o-transcribe` and
+        #   `gpt-4o-mini-transcribe`, the only supported format is `json`. For
+        #   `gpt-4o-transcribe-diarize`, the supported formats are `json`, `text`, and
+        #   `diarized_json`, with `diarized_json` required to receive speaker annotations.
         #
         #   @return [Symbol, OpenAI::Models::AudioResponseFormat, nil]
         optional :response_format, enum: -> { OpenAI::AudioResponseFormat }
@@ -86,13 +110,14 @@ module OpenAI
         #   `response_format` must be set `verbose_json` to use timestamp granularities.
         #   Either or both of these options are supported: `word`, or `segment`. Note: There
         #   is no additional latency for segment timestamps, but generating word timestamps
-        #   incurs additional latency.
+        #   incurs additional latency. This option is not available for
+        #   `gpt-4o-transcribe-diarize`.
         #
         #   @return [Array<Symbol, OpenAI::Models::Audio::TranscriptionCreateParams::TimestampGranularity>, nil]
         optional :timestamp_granularities,
                  -> { OpenAI::Internal::Type::ArrayOf[enum: OpenAI::Audio::TranscriptionCreateParams::TimestampGranularity] }
 
-        # @!method initialize(file:, model:, chunking_strategy: nil, include: nil, language: nil, prompt: nil, response_format: nil, temperature: nil, timestamp_granularities: nil, request_options: {})
+        # @!method initialize(file:, model:, chunking_strategy: nil, include: nil, known_speaker_names: nil, known_speaker_references: nil, language: nil, prompt: nil, response_format: nil, temperature: nil, timestamp_granularities: nil, request_options: {})
         #   Some parameter documentations has been truncated, see
         #   {OpenAI::Models::Audio::TranscriptionCreateParams} for more details.
         #
@@ -104,6 +129,10 @@ module OpenAI
         #
         #   @param include [Array<Symbol, OpenAI::Models::Audio::TranscriptionInclude>] Additional information to include in the transcription response.
         #
+        #   @param known_speaker_names [Array<String>] Optional list of speaker names that correspond to the audio samples provided in
+        #
+        #   @param known_speaker_references [Array<String>] Optional list of audio samples (as [data URLs](https://developer.mozilla.org/en-
+        #
         #   @param language [String] The language of the input audio. Supplying the input language in [ISO-639-1](htt
         #
         #   @param prompt [String] An optional text to guide the model's style or continue a previous audio segment
@@ -117,14 +146,14 @@ module OpenAI
         #   @param request_options [OpenAI::RequestOptions, Hash{Symbol=>Object}]
 
         # ID of the model to use. The options are `gpt-4o-transcribe`,
-        # `gpt-4o-mini-transcribe`, and `whisper-1` (which is powered by our open source
-        # Whisper V2 model).
+        # `gpt-4o-mini-transcribe`, `whisper-1` (which is powered by our open source
+        # Whisper V2 model), and `gpt-4o-transcribe-diarize`.
         module Model
           extend OpenAI::Internal::Type::Union
 
           variant String
 
-          # ID of the model to use. The options are `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`, and `whisper-1` (which is powered by our open source Whisper V2 model).
+          # ID of the model to use. The options are `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`, `whisper-1` (which is powered by our open source Whisper V2 model), and `gpt-4o-transcribe-diarize`.
           variant enum: -> { OpenAI::AudioModel }
 
           # @!method self.variants
@@ -135,6 +164,8 @@ module OpenAI
         # first normalizes loudness and then uses voice activity detection (VAD) to choose
         # boundaries. `server_vad` object can be provided to tweak VAD detection
         # parameters manually. If unset, the audio is transcribed as a single block.
+        # Required when using `gpt-4o-transcribe-diarize` for inputs longer than 30
+        # seconds.
         module ChunkingStrategy
           extend OpenAI::Internal::Type::Union
 

data/lib/openai/models/audio/transcription_create_response.rb

@@ -15,11 +15,14 @@ module OpenAI
         # Represents a transcription response returned by model, based on the provided input.
         variant -> { OpenAI::Audio::Transcription }
 
+        # Represents a diarized transcription response returned by the model, including the combined transcript and speaker-segment annotations.
+        variant -> { OpenAI::Audio::TranscriptionDiarized }
+
         # Represents a verbose json transcription response returned by model, based on the provided input.
         variant -> { OpenAI::Audio::TranscriptionVerbose }
 
         # @!method self.variants
-        #   @return [Array(OpenAI::Models::Audio::Transcription, OpenAI::Models::Audio::TranscriptionVerbose)]
+        #   @return [Array(OpenAI::Models::Audio::Transcription, OpenAI::Models::Audio::TranscriptionDiarized, OpenAI::Models::Audio::TranscriptionVerbose)]
       end
     end
   end

data/lib/openai/models/audio/transcription_diarized.rb

@@ -0,0 +1,160 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Audio
+      class TranscriptionDiarized < OpenAI::Internal::Type::BaseModel
+        # @!attribute duration
+        #   Duration of the input audio in seconds.
+        #
+        #   @return [Float]
+        required :duration, Float
+
+        # @!attribute segments
+        #   Segments of the transcript annotated with timestamps and speaker labels.
+        #
+        #   @return [Array<OpenAI::Models::Audio::TranscriptionDiarizedSegment>]
+        required :segments, -> { OpenAI::Internal::Type::ArrayOf[OpenAI::Audio::TranscriptionDiarizedSegment] }
+
+        # @!attribute task
+        #   The type of task that was run. Always `transcribe`.
+        #
+        #   @return [Symbol, :transcribe]
+        required :task, const: :transcribe
+
+        # @!attribute text
+        #   The concatenated transcript text for the entire audio input.
+        #
+        #   @return [String]
+        required :text, String
+
+        # @!attribute usage
+        #   Token or duration usage statistics for the request.
+        #
+        #   @return [OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens, OpenAI::Models::Audio::TranscriptionDiarized::Usage::Duration, nil]
+        optional :usage, union: -> { OpenAI::Audio::TranscriptionDiarized::Usage }
+
+        # @!method initialize(duration:, segments:, text:, usage: nil, task: :transcribe)
+        #   Represents a diarized transcription response returned by the model, including
+        #   the combined transcript and speaker-segment annotations.
+        #
+        #   @param duration [Float] Duration of the input audio in seconds.
+        #
+        #   @param segments [Array<OpenAI::Models::Audio::TranscriptionDiarizedSegment>] Segments of the transcript annotated with timestamps and speaker labels.
+        #
+        #   @param text [String] The concatenated transcript text for the entire audio input.
+        #
+        #   @param usage [OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens, OpenAI::Models::Audio::TranscriptionDiarized::Usage::Duration] Token or duration usage statistics for the request.
+        #
+        #   @param task [Symbol, :transcribe] The type of task that was run. Always `transcribe`.
+
+        # Token or duration usage statistics for the request.
+        #
+        # @see OpenAI::Models::Audio::TranscriptionDiarized#usage
+        module Usage
+          extend OpenAI::Internal::Type::Union
+
+          discriminator :type
+
+          # Usage statistics for models billed by token usage.
+          variant :tokens, -> { OpenAI::Audio::TranscriptionDiarized::Usage::Tokens }
+
+          # Usage statistics for models billed by audio input duration.
+          variant :duration, -> { OpenAI::Audio::TranscriptionDiarized::Usage::Duration }
+
+          class Tokens < OpenAI::Internal::Type::BaseModel
+            # @!attribute input_tokens
+            #   Number of input tokens billed for this request.
+            #
+            #   @return [Integer]
+            required :input_tokens, Integer
+
+            # @!attribute output_tokens
+            #   Number of output tokens generated.
+            #
+            #   @return [Integer]
+            required :output_tokens, Integer
+
+            # @!attribute total_tokens
+            #   Total number of tokens used (input + output).
+            #
+            #   @return [Integer]
+            required :total_tokens, Integer
+
+            # @!attribute type
+            #   The type of the usage object. Always `tokens` for this variant.
+            #
+            #   @return [Symbol, :tokens]
+            required :type, const: :tokens
+
+            # @!attribute input_token_details
+            #   Details about the input tokens billed for this request.
+            #
+            #   @return [OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens::InputTokenDetails, nil]
+            optional :input_token_details,
+                     -> { OpenAI::Audio::TranscriptionDiarized::Usage::Tokens::InputTokenDetails }
+
+            # @!method initialize(input_tokens:, output_tokens:, total_tokens:, input_token_details: nil, type: :tokens)
+            #   Usage statistics for models billed by token usage.
+            #
+            #   @param input_tokens [Integer] Number of input tokens billed for this request.
+            #
+            #   @param output_tokens [Integer] Number of output tokens generated.
+            #
+            #   @param total_tokens [Integer] Total number of tokens used (input + output).
+            #
+            #   @param input_token_details [OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens::InputTokenDetails] Details about the input tokens billed for this request.
+            #
+            #   @param type [Symbol, :tokens] The type of the usage object. Always `tokens` for this variant.
+
+            # @see OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens#input_token_details
+            class InputTokenDetails < OpenAI::Internal::Type::BaseModel
+              # @!attribute audio_tokens
+              #   Number of audio tokens billed for this request.
+              #
+              #   @return [Integer, nil]
+              optional :audio_tokens, Integer
+
+              # @!attribute text_tokens
+              #   Number of text tokens billed for this request.
+              #
+              #   @return [Integer, nil]
+              optional :text_tokens, Integer
+
+              # @!method initialize(audio_tokens: nil, text_tokens: nil)
+              #   Details about the input tokens billed for this request.
+              #
+              #   @param audio_tokens [Integer] Number of audio tokens billed for this request.
+              #
+              #   @param text_tokens [Integer] Number of text tokens billed for this request.
+            end
+          end
+
+          class Duration < OpenAI::Internal::Type::BaseModel
+            # @!attribute seconds
+            #   Duration of the input audio in seconds.
+            #
+            #   @return [Float]
+            required :seconds, Float
+
+            # @!attribute type
+            #   The type of the usage object. Always `duration` for this variant.
+            #
+            #   @return [Symbol, :duration]
+            required :type, const: :duration
+
+            # @!method initialize(seconds:, type: :duration)
+            #   Usage statistics for models billed by audio input duration.
+            #
+            #   @param seconds [Float] Duration of the input audio in seconds.
+            #
+            #   @param type [Symbol, :duration] The type of the usage object. Always `duration` for this variant.
+          end
+
+          # @!method self.variants
+          #   @return [Array(OpenAI::Models::Audio::TranscriptionDiarized::Usage::Tokens, OpenAI::Models::Audio::TranscriptionDiarized::Usage::Duration)]
+        end
+      end
+    end
+  end
+end

data/lib/openai/models/audio/transcription_diarized_segment.rb

@@ -0,0 +1,65 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Audio
+      class TranscriptionDiarizedSegment < OpenAI::Internal::Type::BaseModel
+        # @!attribute id
+        #   Unique identifier for the segment.
+        #
+        #   @return [String]
+        required :id, String
+
+        # @!attribute end_
+        #   End timestamp of the segment in seconds.
+        #
+        #   @return [Float]
+        required :end_, Float, api_name: :end
+
+        # @!attribute speaker
+        #   Speaker label for this segment. When known speakers are provided, the label
+        #   matches `known_speaker_names[]`. Otherwise speakers are labeled sequentially
+        #   using capital letters (`A`, `B`, ...).
+        #
+        #   @return [String]
+        required :speaker, String
+
+        # @!attribute start
+        #   Start timestamp of the segment in seconds.
+        #
+        #   @return [Float]
+        required :start, Float
+
+        # @!attribute text
+        #   Transcript text for this segment.
+        #
+        #   @return [String]
+        required :text, String
+
+        # @!attribute type
+        #   The type of the segment. Always `transcript.text.segment`.
+        #
+        #   @return [Symbol, :"transcript.text.segment"]
+        required :type, const: :"transcript.text.segment"
+
+        # @!method initialize(id:, end_:, speaker:, start:, text:, type: :"transcript.text.segment")
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Audio::TranscriptionDiarizedSegment} for more details.
+        #
+        #   A segment of diarized transcript text with speaker metadata.
+        #
+        #   @param id [String] Unique identifier for the segment.
+        #
+        #   @param end_ [Float] End timestamp of the segment in seconds.
+        #
+        #   @param speaker [String] Speaker label for this segment. When known speakers are provided, the label matc
+        #
+        #   @param start [Float] Start timestamp of the segment in seconds.
+        #
+        #   @param text [String] Transcript text for this segment.
+        #
+        #   @param type [Symbol, :"transcript.text.segment"] The type of the segment. Always `transcript.text.segment`.
+      end
+    end
+  end
+end

data/lib/openai/models/audio/transcription_stream_event.rb

@@ -3,15 +3,18 @@
 module OpenAI
   module Models
     module Audio
-      # Emitted when there is an additional text delta. This is also the first event
-      # emitted when the transcription starts. Only emitted when you
+      # Emitted when a diarized transcription returns a completed segment with speaker
+      # information. Only emitted when you
       # [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription)
-      # with the `Stream` parameter set to `true`.
+      # with `stream` set to `true` and `response_format` set to `diarized_json`.
       module TranscriptionStreamEvent
         extend OpenAI::Internal::Type::Union
 
         discriminator :type
 
+        # Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with `stream` set to `true` and `response_format` set to `diarized_json`.
+        variant :"transcript.text.segment", -> { OpenAI::Audio::TranscriptionTextSegmentEvent }
+
         # Emitted when there is an additional text delta. This is also the first event emitted when the transcription starts. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `Stream` parameter set to `true`.
         variant :"transcript.text.delta", -> { OpenAI::Audio::TranscriptionTextDeltaEvent }
 
@@ -19,7 +22,7 @@ module OpenAI
         variant :"transcript.text.done", -> { OpenAI::Audio::TranscriptionTextDoneEvent }
 
         # @!method self.variants
-        #   @return [Array(OpenAI::Models::Audio::TranscriptionTextDeltaEvent, OpenAI::Models::Audio::TranscriptionTextDoneEvent)]
+        #   @return [Array(OpenAI::Models::Audio::TranscriptionTextSegmentEvent, OpenAI::Models::Audio::TranscriptionTextDeltaEvent, OpenAI::Models::Audio::TranscriptionTextDoneEvent)]
       end
     end
   end

data/lib/openai/models/audio/transcription_text_delta_event.rb

@@ -25,7 +25,14 @@ module OpenAI
         optional :logprobs,
                  -> { OpenAI::Internal::Type::ArrayOf[OpenAI::Audio::TranscriptionTextDeltaEvent::Logprob] }
 
-        # @!method initialize(delta:, logprobs: nil, type: :"transcript.text.delta")
+        # @!attribute segment_id
+        #   Identifier of the diarized segment that this delta belongs to. Only present when
+        #   using `gpt-4o-transcribe-diarize`.
+        #
+        #   @return [String, nil]
+        optional :segment_id, String
+
+        # @!method initialize(delta:, logprobs: nil, segment_id: nil, type: :"transcript.text.delta")
         #   Some parameter documentations has been truncated, see
         #   {OpenAI::Models::Audio::TranscriptionTextDeltaEvent} for more details.
         #
@@ -38,6 +45,8 @@ module OpenAI
         #
         #   @param logprobs [Array<OpenAI::Models::Audio::TranscriptionTextDeltaEvent::Logprob>] The log probabilities of the delta. Only included if you [create a transcription
         #
+        #   @param segment_id [String] Identifier of the diarized segment that this delta belongs to. Only present when
+        #
         #   @param type [Symbol, :"transcript.text.delta"] The type of the event. Always `transcript.text.delta`.
 
         class Logprob < OpenAI::Internal::Type::BaseModel

data/lib/openai/models/audio/transcription_text_segment_event.rb

@@ -0,0 +1,63 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Audio
+      class TranscriptionTextSegmentEvent < OpenAI::Internal::Type::BaseModel
+        # @!attribute id
+        #   Unique identifier for the segment.
+        #
+        #   @return [String]
+        required :id, String
+
+        # @!attribute end_
+        #   End timestamp of the segment in seconds.
+        #
+        #   @return [Float]
+        required :end_, Float, api_name: :end
+
+        # @!attribute speaker
+        #   Speaker label for this segment.
+        #
+        #   @return [String]
+        required :speaker, String
+
+        # @!attribute start
+        #   Start timestamp of the segment in seconds.
+        #
+        #   @return [Float]
+        required :start, Float
+
+        # @!attribute text
+        #   Transcript text for this segment.
+        #
+        #   @return [String]
+        required :text, String
+
+        # @!attribute type
+        #   The type of the event. Always `transcript.text.segment`.
+        #
+        #   @return [Symbol, :"transcript.text.segment"]
+        required :type, const: :"transcript.text.segment"
+
+        # @!method initialize(id:, end_:, speaker:, start:, text:, type: :"transcript.text.segment")
+        #   Emitted when a diarized transcription returns a completed segment with speaker
+        #   information. Only emitted when you
+        #   [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription)
+        #   with `stream` set to `true` and `response_format` set to `diarized_json`.
+        #
+        #   @param id [String] Unique identifier for the segment.
+        #
+        #   @param end_ [Float] End timestamp of the segment in seconds.
+        #
+        #   @param speaker [String] Speaker label for this segment.
+        #
+        #   @param start [Float] Start timestamp of the segment in seconds.
+        #
+        #   @param text [String] Transcript text for this segment.
+        #
+        #   @param type [Symbol, :"transcript.text.segment"] The type of the event. Always `transcript.text.segment`.
+      end
+    end
+  end
+end

data/lib/openai/models/audio_model.rb

@@ -8,6 +8,7 @@ module OpenAI
       WHISPER_1 = :"whisper-1"
       GPT_4O_TRANSCRIBE = :"gpt-4o-transcribe"
       GPT_4O_MINI_TRANSCRIBE = :"gpt-4o-mini-transcribe"
+      GPT_4O_TRANSCRIBE_DIARIZE = :"gpt-4o-transcribe-diarize"
 
       # @!method self.values
       #   @return [Array<Symbol>]

data/lib/openai/models/audio_response_format.rb

@@ -3,8 +3,10 @@
 module OpenAI
   module Models
     # The format of the output, in one of these options: `json`, `text`, `srt`,
-    # `verbose_json`, or `vtt`. For `gpt-4o-transcribe` and `gpt-4o-mini-transcribe`,
-    # the only supported format is `json`.
+    # `verbose_json`, `vtt`, or `diarized_json`. For `gpt-4o-transcribe` and
+    # `gpt-4o-mini-transcribe`, the only supported format is `json`. For
+    # `gpt-4o-transcribe-diarize`, the supported formats are `json`, `text`, and
+    # `diarized_json`, with `diarized_json` required to receive speaker annotations.
     module AudioResponseFormat
       extend OpenAI::Internal::Type::Enum
 
@@ -13,6 +15,7 @@ module OpenAI
       SRT = :srt
       VERBOSE_JSON = :verbose_json
       VTT = :vtt
+      DIARIZED_JSON = :diarized_json
 
       # @!method self.values
       #   @return [Array<Symbol>]

data/lib/openai/models/realtime/audio_transcription.rb

@@ -14,7 +14,8 @@ module OpenAI
 
         # @!attribute model
         #   The model to use for transcription. Current options are `whisper-1`,
-        #   `gpt-4o-transcribe-latest`, `gpt-4o-mini-transcribe`, and `gpt-4o-transcribe`.
+        #   `gpt-4o-mini-transcribe`, `gpt-4o-transcribe`, and `gpt-4o-transcribe-diarize`.
+        #   Use `gpt-4o-transcribe-diarize` when you need diarization with speaker labels.
         #
         #   @return [Symbol, OpenAI::Models::Realtime::AudioTranscription::Model, nil]
         optional :model, enum: -> { OpenAI::Realtime::AudioTranscription::Model }
@@ -23,8 +24,8 @@ module OpenAI
         #   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](https://platform.openai.com/docs/guides/speech-to-text#prompting).
-        #   For `gpt-4o-transcribe` models, the prompt is a free text string, for example
-        #   "expect words related to technology".
+        #   For `gpt-4o-transcribe` models (excluding `gpt-4o-transcribe-diarize`), the
+        #   prompt is a free text string, for example "expect words related to technology".
         #
         #   @return [String, nil]
         optional :prompt, String
@@ -35,21 +36,22 @@ module OpenAI
         #
         #   @param language [String] The language of the input audio. Supplying the input language in
         #
-        #   @param model [Symbol, OpenAI::Models::Realtime::AudioTranscription::Model] The model to use for transcription. Current options are `whisper-1`, `gpt-4o-tra
+        #   @param model [Symbol, OpenAI::Models::Realtime::AudioTranscription::Model] The model to use for transcription. Current options are `whisper-1`, `gpt-4o-min
         #
         #   @param prompt [String] An optional text to guide the model's style or continue a previous audio
 
         # The model to use for transcription. Current options are `whisper-1`,
-        # `gpt-4o-transcribe-latest`, `gpt-4o-mini-transcribe`, and `gpt-4o-transcribe`.
+        # `gpt-4o-mini-transcribe`, `gpt-4o-transcribe`, and `gpt-4o-transcribe-diarize`.
+        # Use `gpt-4o-transcribe-diarize` when you need diarization with speaker labels.
         #
         # @see OpenAI::Models::Realtime::AudioTranscription#model
         module Model
           extend OpenAI::Internal::Type::Enum
 
           WHISPER_1 = :"whisper-1"
-          GPT_4O_TRANSCRIBE_LATEST = :"gpt-4o-transcribe-latest"
           GPT_4O_MINI_TRANSCRIBE = :"gpt-4o-mini-transcribe"
           GPT_4O_TRANSCRIBE = :"gpt-4o-transcribe"
+          GPT_4O_TRANSCRIBE_DIARIZE = :"gpt-4o-transcribe-diarize"
 
           # @!method self.values
           #   @return [Array<Symbol>]