openai 0.61.0 → 0.62.0

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

@@ -16,6 +16,9 @@ module OpenAI
       # 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.
       module RealtimeTranscriptionSessionAudioInputTurnDetection
         extend OpenAI::Internal::Type::Union
 

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

@@ -91,7 +91,6 @@ module OpenAI
                      -> { OpenAI::Realtime::RealtimeTranscriptionSessionCreateResponse::Audio::Input::NoiseReduction }
 
             # @!attribute transcription
-            #   Configuration of the transcription model.
             #
             #   @return [OpenAI::Models::Realtime::AudioTranscription, nil]
             optional :transcription, -> { OpenAI::Realtime::AudioTranscription }
@@ -99,10 +98,15 @@ module OpenAI
             # @!attribute turn_detection
             #   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.
+            #   volume and respond at the end of user speech. For `gpt-realtime-whisper`, this
+            #   must be `null`; VAD is not supported.
             #
             #   @return [OpenAI::Models::Realtime::RealtimeTranscriptionSessionTurnDetection, nil]
-            optional :turn_detection, -> { OpenAI::Realtime::RealtimeTranscriptionSessionTurnDetection }
+            optional :turn_detection,
+                     -> {
+                       OpenAI::Realtime::RealtimeTranscriptionSessionTurnDetection
+                     },
+                     nil?: true
 
             # @!method initialize(format_: nil, noise_reduction: nil, transcription: nil, turn_detection: nil)
             #   Some parameter documentations has been truncated, see
@@ -113,9 +117,9 @@ module OpenAI
             #
             #   @param noise_reduction [OpenAI::Models::Realtime::RealtimeTranscriptionSessionCreateResponse::Audio::Input::NoiseReduction] Configuration for input audio noise reduction.
             #
-            #   @param transcription [OpenAI::Models::Realtime::AudioTranscription] Configuration of the transcription model.
+            #   @param transcription [OpenAI::Models::Realtime::AudioTranscription]
             #
-            #   @param turn_detection [OpenAI::Models::Realtime::RealtimeTranscriptionSessionTurnDetection] Configuration for turn detection. Can be set to `null` to turn off. Server
+            #   @param turn_detection [OpenAI::Models::Realtime::RealtimeTranscriptionSessionTurnDetection, nil] Configuration for turn detection. Can be set to `null` to turn off. Server
 
             # @see OpenAI::Models::Realtime::RealtimeTranscriptionSessionCreateResponse::Audio::Input#noise_reduction
             class NoiseReduction < OpenAI::Internal::Type::BaseModel

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

@@ -40,7 +40,8 @@ module OpenAI
         #
         #   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.
+        #   volume and respond at the end of user speech. For `gpt-realtime-whisper`, this
+        #   must be `null`; VAD is not supported.
         #
         #   @param prefix_padding_ms [Integer] Amount of audio to include before the VAD detected speech (in
         #

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

@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      # A Realtime translation client event.
+      module RealtimeTranslationClientEvent
+        extend OpenAI::Internal::Type::Union
+
+        discriminator :type
+
+        # Send this event to update the translation session configuration. Translation
+        # sessions support updates to `audio.output.language`, `audio.input.transcription`,
+        # and `audio.input.noise_reduction`.
+        variant :"session.update", -> { OpenAI::Realtime::RealtimeTranslationSessionUpdateEvent }
+
+        # Send this event to append audio bytes to the translation session input audio buffer.
+        #
+        # WebSocket translation sessions accept base64-encoded 24 kHz PCM16 mono
+        # little-endian raw audio bytes. Unsupported websocket audio formats return a
+        # validation error because lower-quality audio materially degrades translation
+        # quality.
+        #
+        # Translation consumes 200 ms engine frames. For best realtime behavior, append
+        # audio in 200 ms chunks. If a chunk is shorter, the server buffers it until it
+        # has enough audio for one frame. If a chunk is longer, the server splits it into
+        # 200 ms frames and enqueues them back-to-back.
+        #
+        # Keep appending silence while the session is active. If a client stops sending
+        # audio and later resumes, model time treats the resumed audio as contiguous with
+        # the previous audio rather than as a real-world pause.
+        variant :"session.input_audio_buffer.append",
+                -> { OpenAI::Realtime::RealtimeTranslationInputAudioBufferAppendEvent }
+
+        # Gracefully close the realtime translation session. The server flushes pending
+        # input audio and emits any remaining translated output before closing the
+        # session.
+        variant :"session.close", -> { OpenAI::Realtime::RealtimeTranslationSessionCloseEvent }
+
+        # @!method self.variants
+        #   @return [Array(OpenAI::Models::Realtime::RealtimeTranslationSessionUpdateEvent, OpenAI::Models::Realtime::RealtimeTranslationInputAudioBufferAppendEvent, OpenAI::Models::Realtime::RealtimeTranslationSessionCloseEvent)]
+      end
+    end
+  end
+end

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

@@ -0,0 +1,85 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationClientSecretCreateRequest < OpenAI::Internal::Type::BaseModel
+        # @!attribute session
+        #   Realtime translation session configuration. Translation sessions stream source
+        #   audio in and translated audio plus transcript deltas out continuously.
+        #
+        #   @return [OpenAI::Models::Realtime::RealtimeTranslationSessionCreateRequest]
+        required :session, -> { OpenAI::Realtime::RealtimeTranslationSessionCreateRequest }
+
+        # @!attribute expires_after
+        #   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.
+        #
+        #   @return [OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter, nil]
+        optional :expires_after,
+                 -> { OpenAI::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter }
+
+        # @!method initialize(session:, expires_after: nil)
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest} for
+        #   more details.
+        #
+        #   Create a translation session and client secret for the Realtime API.
+        #
+        #   @param session [OpenAI::Models::Realtime::RealtimeTranslationSessionCreateRequest] Realtime translation session configuration. Translation sessions stream source
+        #
+        #   @param expires_after [OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter] Configuration for the client secret expiration. Expiration refers to the time af
+
+        # @see OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest#expires_after
+        class ExpiresAfter < OpenAI::Internal::Type::BaseModel
+          # @!attribute anchor
+          #   The anchor point for the client secret expiration, meaning that `seconds` will
+          #   be added to the `created_at` time of the client secret to produce an expiration
+          #   timestamp. Only `created_at` is currently supported.
+          #
+          #   @return [Symbol, OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter::Anchor, nil]
+          optional :anchor,
+                   enum: -> { OpenAI::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter::Anchor }
+
+          # @!attribute seconds
+          #   The number of seconds from the anchor point to the expiration. Select a value
+          #   between `10` and `7200` (2 hours). This default to 600 seconds (10 minutes) if
+          #   not specified.
+          #
+          #   @return [Integer, nil]
+          optional :seconds, Integer
+
+          # @!method initialize(anchor: nil, seconds: nil)
+          #   Some parameter documentations has been truncated, see
+          #   {OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter}
+          #   for more details.
+          #
+          #   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.
+          #
+          #   @param anchor [Symbol, OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter::Anchor] The anchor point for the client secret expiration, meaning that `seconds` will b
+          #
+          #   @param seconds [Integer] The number of seconds from the anchor point to the expiration. Select a value be
+
+          # The anchor point for the client secret expiration, meaning that `seconds` will
+          # be added to the `created_at` time of the client secret to produce an expiration
+          # timestamp. Only `created_at` is currently supported.
+          #
+          # @see OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateRequest::ExpiresAfter#anchor
+          module Anchor
+            extend OpenAI::Internal::Type::Enum
+
+            CREATED_AT = :created_at
+
+            # @!method self.values
+            #   @return [Array<Symbol>]
+          end
+        end
+      end
+    end
+  end
+end

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

@@ -0,0 +1,42 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationClientSecretCreateResponse < OpenAI::Internal::Type::BaseModel
+        # @!attribute expires_at
+        #   Expiration timestamp for the client secret, in seconds since epoch.
+        #
+        #   @return [Integer]
+        required :expires_at, Integer
+
+        # @!attribute session
+        #   A Realtime translation session. Translation sessions continuously translate
+        #   input audio into the configured output language.
+        #
+        #   @return [OpenAI::Models::Realtime::RealtimeTranslationSession]
+        required :session, -> { OpenAI::Realtime::RealtimeTranslationSession }
+
+        # @!attribute value
+        #   The generated client secret value.
+        #
+        #   @return [String]
+        required :value, String
+
+        # @!method initialize(expires_at:, session:, value:)
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Realtime::RealtimeTranslationClientSecretCreateResponse} for
+        #   more details.
+        #
+        #   Response from creating a translation session and client secret for the Realtime
+        #   API.
+        #
+        #   @param expires_at [Integer] Expiration timestamp for the client secret, in seconds since epoch.
+        #
+        #   @param session [OpenAI::Models::Realtime::RealtimeTranslationSession] A Realtime translation session. Translation sessions continuously translate inpu
+        #
+        #   @param value [String] The generated client secret value.
+      end
+    end
+  end
+end

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

@@ -0,0 +1,51 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationInputAudioBufferAppendEvent < OpenAI::Internal::Type::BaseModel
+        # @!attribute audio
+        #   Base64-encoded 24 kHz PCM16 mono audio bytes.
+        #
+        #   @return [String]
+        required :audio, String
+
+        # @!attribute type
+        #   The event type, must be `session.input_audio_buffer.append`.
+        #
+        #   @return [Symbol, :"session.input_audio_buffer.append"]
+        required :type, const: :"session.input_audio_buffer.append"
+
+        # @!attribute event_id
+        #   Optional client-generated ID used to identify this event.
+        #
+        #   @return [String, nil]
+        optional :event_id, String
+
+        # @!method initialize(audio:, event_id: nil, type: :"session.input_audio_buffer.append")
+        #   Send this event to append audio bytes to the translation session input audio
+        #   buffer.
+        #
+        #   WebSocket translation sessions accept base64-encoded 24 kHz PCM16 mono
+        #   little-endian raw audio bytes. Unsupported websocket audio formats return a
+        #   validation error because lower-quality audio materially degrades translation
+        #   quality.
+        #
+        #   Translation consumes 200 ms engine frames. For best realtime behavior, append
+        #   audio in 200 ms chunks. If a chunk is shorter, the server buffers it until it
+        #   has enough audio for one frame. If a chunk is longer, the server splits it into
+        #   200 ms frames and enqueues them back-to-back.
+        #
+        #   Keep appending silence while the session is active. If a client stops sending
+        #   audio and later resumes, model time treats the resumed audio as contiguous with
+        #   the previous audio rather than as a real-world pause.
+        #
+        #   @param audio [String] Base64-encoded 24 kHz PCM16 mono audio bytes.
+        #
+        #   @param event_id [String] Optional client-generated ID used to identify this event.
+        #
+        #   @param type [Symbol, :"session.input_audio_buffer.append"] The event type, must be `session.input_audio_buffer.append`.
+      end
+    end
+  end
+end

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

@@ -0,0 +1,55 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationInputTranscriptDeltaEvent < OpenAI::Internal::Type::BaseModel
+        # @!attribute delta
+        #   Append-only source-language transcript text.
+        #
+        #   @return [String]
+        required :delta, String
+
+        # @!attribute event_id
+        #   The unique ID of the server event.
+        #
+        #   @return [String]
+        required :event_id, String
+
+        # @!attribute type
+        #   The event type, must be `session.input_transcript.delta`.
+        #
+        #   @return [Symbol, :"session.input_transcript.delta"]
+        required :type, const: :"session.input_transcript.delta"
+
+        # @!attribute elapsed_ms
+        #   Timing metadata for stream alignment, derived from the translation frame when
+        #   available. It advances in 200 ms increments, but multiple transcript deltas may
+        #   share the same `elapsed_ms`. Treat it as alignment metadata, not a unique
+        #   transcript-delta identifier.
+        #
+        #   @return [Integer, nil]
+        optional :elapsed_ms, Integer, nil?: true
+
+        # @!method initialize(delta:, event_id:, elapsed_ms: nil, type: :"session.input_transcript.delta")
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Realtime::RealtimeTranslationInputTranscriptDeltaEvent} for
+        #   more details.
+        #
+        #   Returned when optional source-language transcript text is available. This event
+        #   is emitted only when `audio.input.transcription` is configured.
+        #
+        #   Transcript deltas are append-only text fragments. Clients should not insert
+        #   unconditional spaces between deltas.
+        #
+        #   @param delta [String] Append-only source-language transcript text.
+        #
+        #   @param event_id [String] The unique ID of the server event.
+        #
+        #   @param elapsed_ms [Integer, nil] Timing metadata for stream alignment, derived from the translation frame
+        #
+        #   @param type [Symbol, :"session.input_transcript.delta"] The event type, must be `session.input_transcript.delta`.
+      end
+    end
+  end
+end

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

@@ -0,0 +1,89 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationOutputAudioDeltaEvent < OpenAI::Internal::Type::BaseModel
+        # @!attribute delta
+        #   Base64-encoded translated audio data.
+        #
+        #   @return [String]
+        required :delta, String
+
+        # @!attribute event_id
+        #   The unique ID of the server event.
+        #
+        #   @return [String]
+        required :event_id, String
+
+        # @!attribute type
+        #   The event type, must be `session.output_audio.delta`.
+        #
+        #   @return [Symbol, :"session.output_audio.delta"]
+        required :type, const: :"session.output_audio.delta"
+
+        # @!attribute channels
+        #   Number of audio channels.
+        #
+        #   @return [Integer, nil]
+        optional :channels, Integer
+
+        # @!attribute elapsed_ms
+        #   Timing metadata for stream alignment, derived from the translation frame when
+        #   available. Treat `elapsed_ms` as alignment metadata, not a unique event
+        #   identifier.
+        #
+        #   @return [Integer, nil]
+        optional :elapsed_ms, Integer, nil?: true
+
+        # @!attribute format_
+        #   Audio encoding for `delta`.
+        #
+        #   @return [Symbol, OpenAI::Models::Realtime::RealtimeTranslationOutputAudioDeltaEvent::Format, nil]
+        optional :format_,
+                 enum: -> { OpenAI::Realtime::RealtimeTranslationOutputAudioDeltaEvent::Format },
+                 api_name: :format
+
+        # @!attribute sample_rate
+        #   Sample rate of the audio delta.
+        #
+        #   @return [Integer, nil]
+        optional :sample_rate, Integer
+
+        # @!method initialize(delta:, event_id:, channels: nil, elapsed_ms: nil, format_: nil, sample_rate: nil, type: :"session.output_audio.delta")
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Realtime::RealtimeTranslationOutputAudioDeltaEvent} for more
+        #   details.
+        #
+        #   Returned when translated output audio is available. Output audio deltas are 200
+        #   ms frames of PCM16 audio.
+        #
+        #   @param delta [String] Base64-encoded translated audio data.
+        #
+        #   @param event_id [String] The unique ID of the server event.
+        #
+        #   @param channels [Integer] Number of audio channels.
+        #
+        #   @param elapsed_ms [Integer, nil] Timing metadata for stream alignment, derived from the translation frame
+        #
+        #   @param format_ [Symbol, OpenAI::Models::Realtime::RealtimeTranslationOutputAudioDeltaEvent::Format] Audio encoding for `delta`.
+        #
+        #   @param sample_rate [Integer] Sample rate of the audio delta.
+        #
+        #   @param type [Symbol, :"session.output_audio.delta"] The event type, must be `session.output_audio.delta`.
+
+        # Audio encoding for `delta`.
+        #
+        # @see OpenAI::Models::Realtime::RealtimeTranslationOutputAudioDeltaEvent#format_
+        module Format
+          extend OpenAI::Internal::Type::Enum
+
+          PCM16 = :pcm16
+
+          # @!method self.values
+          #   @return [Array<Symbol>]
+        end
+      end
+    end
+  end
+end

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

@@ -0,0 +1,54 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      class RealtimeTranslationOutputTranscriptDeltaEvent < OpenAI::Internal::Type::BaseModel
+        # @!attribute delta
+        #   Append-only transcript text for the translated output audio.
+        #
+        #   @return [String]
+        required :delta, String
+
+        # @!attribute event_id
+        #   The unique ID of the server event.
+        #
+        #   @return [String]
+        required :event_id, String
+
+        # @!attribute type
+        #   The event type, must be `session.output_transcript.delta`.
+        #
+        #   @return [Symbol, :"session.output_transcript.delta"]
+        required :type, const: :"session.output_transcript.delta"
+
+        # @!attribute elapsed_ms
+        #   Timing metadata for stream alignment, derived from the translation frame when
+        #   available. It advances in 200 ms increments, but multiple transcript deltas may
+        #   share the same `elapsed_ms`. Treat it as alignment metadata, not a unique
+        #   transcript-delta identifier.
+        #
+        #   @return [Integer, nil]
+        optional :elapsed_ms, Integer, nil?: true
+
+        # @!method initialize(delta:, event_id:, elapsed_ms: nil, type: :"session.output_transcript.delta")
+        #   Some parameter documentations has been truncated, see
+        #   {OpenAI::Models::Realtime::RealtimeTranslationOutputTranscriptDeltaEvent} for
+        #   more details.
+        #
+        #   Returned when translated transcript text is available.
+        #
+        #   Transcript deltas are append-only text fragments. Clients should not insert
+        #   unconditional spaces between deltas.
+        #
+        #   @param delta [String] Append-only transcript text for the translated output audio.
+        #
+        #   @param event_id [String] The unique ID of the server event.
+        #
+        #   @param elapsed_ms [Integer, nil] Timing metadata for stream alignment, derived from the translation frame
+        #
+        #   @param type [Symbol, :"session.output_transcript.delta"] The event type, must be `session.output_transcript.delta`.
+      end
+    end
+  end
+end

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

@@ -0,0 +1,53 @@
+# frozen_string_literal: true
+
+module OpenAI
+  module Models
+    module Realtime
+      # A Realtime translation server event.
+      module RealtimeTranslationServerEvent
+        extend OpenAI::Internal::Type::Union
+
+        discriminator :type
+
+        # Returned when an error occurs, which could be a client problem or a server
+        # problem. Most errors are recoverable and the session will stay open, we
+        # recommend to implementors to monitor and log error messages by default.
+        variant :error, -> { OpenAI::Realtime::RealtimeErrorEvent }
+
+        # Returned when a translation session is created. Emitted automatically when a
+        # new connection is established as the first server event. This event contains
+        # the default translation session configuration.
+        variant :"session.created", -> { OpenAI::Realtime::RealtimeTranslationSessionCreatedEvent }
+
+        # Returned when a translation session is updated with a `session.update` event,
+        # unless there is an error.
+        variant :"session.updated", -> { OpenAI::Realtime::RealtimeTranslationSessionUpdatedEvent }
+
+        # Returned when a realtime translation session is closed.
+        variant :"session.closed", -> { OpenAI::Realtime::RealtimeTranslationSessionClosedEvent }
+
+        # Returned when optional source-language transcript text is available. This event
+        # is emitted only when `audio.input.transcription` is configured.
+        #
+        # Transcript deltas are append-only text fragments. Clients should not insert
+        # unconditional spaces between deltas.
+        variant :"session.input_transcript.delta",
+                -> { OpenAI::Realtime::RealtimeTranslationInputTranscriptDeltaEvent }
+
+        # Returned when translated transcript text is available.
+        #
+        # Transcript deltas are append-only text fragments. Clients should not insert
+        # unconditional spaces between deltas.
+        variant :"session.output_transcript.delta",
+                -> { OpenAI::Realtime::RealtimeTranslationOutputTranscriptDeltaEvent }
+
+        # Returned when translated output audio is available. Output audio deltas are
+        # 200 ms frames of PCM16 audio.
+        variant :"session.output_audio.delta", -> { OpenAI::Realtime::RealtimeTranslationOutputAudioDeltaEvent }
+
+        # @!method self.variants
+        #   @return [Array(OpenAI::Models::Realtime::RealtimeErrorEvent, OpenAI::Models::Realtime::RealtimeTranslationSessionCreatedEvent, OpenAI::Models::Realtime::RealtimeTranslationSessionUpdatedEvent, OpenAI::Models::Realtime::RealtimeTranslationSessionClosedEvent, OpenAI::Models::Realtime::RealtimeTranslationInputTranscriptDeltaEvent, OpenAI::Models::Realtime::RealtimeTranslationOutputTranscriptDeltaEvent, OpenAI::Models::Realtime::RealtimeTranslationOutputAudioDeltaEvent)]
+      end
+    end
+  end
+end