telnyx 5.167.0 → 5.169.0

data/rbi/telnyx/models/ai/audio_transcribe_response.rbi

@@ -17,8 +17,9 @@ module Telnyx
         attr_accessor :text
 
         # The duration of the audio file in seconds. Returned by
-        # `distil-whisper/distil-large-v2` and `deepgram/nova-3` when `response_format` is
-        # `verbose_json`. Not returned by `openai/whisper-large-v3-turbo`.
+        # `distil-whisper/distil-large-v2` and the `deepgram/*` models when
+        # `response_format` is `verbose_json`. Not returned by
+        # `openai/whisper-large-v3-turbo`.
         sig { returns(T.nilable(Float)) }
         attr_reader :duration
 
@@ -26,8 +27,9 @@ module Telnyx
         attr_writer :duration
 
         # Segments of the transcribed text and their corresponding details. Returned by
-        # `distil-whisper/distil-large-v2` when `response_format` is `verbose_json`. Not
-        # returned by `openai/whisper-large-v3-turbo`.
+        # `distil-whisper/distil-large-v2` and the `deepgram/*` models when
+        # `response_format` is `verbose_json`; Deepgram segments also carry nested `words`
+        # and `speakers`. Not returned by `openai/whisper-large-v3-turbo`.
         sig do
           returns(
             T.nilable(
@@ -47,32 +49,28 @@ module Telnyx
         end
         attr_writer :segments
 
-        # Word-level timestamps and optional speaker labels. Only returned by
-        # `deepgram/nova-3` when word-level output is enabled via `model_config`.
+        # Word-level timestamps and optional speaker labels. Only returned by the
+        # `deepgram/*` models when word-level output is enabled via `model_config`.
         sig do
           returns(
-            T.nilable(
-              T::Array[Telnyx::Models::AI::AudioTranscribeResponse::Word]
-            )
+            T.nilable(T::Array[Telnyx::AI::AudioTranscriptionResponseWord])
           )
         end
         attr_reader :words
 
         sig do
           params(
-            words:
-              T::Array[
-                Telnyx::Models::AI::AudioTranscribeResponse::Word::OrHash
-              ]
+            words: T::Array[Telnyx::AI::AudioTranscriptionResponseWord::OrHash]
           ).void
         end
         attr_writer :words
 
         # Response fields vary by model. `distil-whisper/distil-large-v2` returns `text`,
         # `duration`, and `segments` in `verbose_json` mode.
-        # `openai/whisper-large-v3-turbo` returns `text` only. `deepgram/nova-3` returns
-        # `text` and, depending on `model_config`, may include `words` with per-word
-        # timestamps and speaker labels.
+        # `openai/whisper-large-v3-turbo` returns `text` only. The `deepgram/*` models
+        # return `text` and, depending on `model_config`, may include `words` with
+        # per-word timestamps and speaker labels. The Parakeet models
+        # (`nvidia/parakeet-v3`, `omi-health/omi-med-stt-v1`) return `text` only.
         sig do
           params(
             text: String,
@@ -81,25 +79,24 @@ module Telnyx
               T::Array[
                 Telnyx::Models::AI::AudioTranscribeResponse::Segment::OrHash
               ],
-            words:
-              T::Array[
-                Telnyx::Models::AI::AudioTranscribeResponse::Word::OrHash
-              ]
+            words: T::Array[Telnyx::AI::AudioTranscriptionResponseWord::OrHash]
           ).returns(T.attached_class)
         end
         def self.new(
           # The transcribed text for the audio file.
           text:,
           # The duration of the audio file in seconds. Returned by
-          # `distil-whisper/distil-large-v2` and `deepgram/nova-3` when `response_format` is
-          # `verbose_json`. Not returned by `openai/whisper-large-v3-turbo`.
+          # `distil-whisper/distil-large-v2` and the `deepgram/*` models when
+          # `response_format` is `verbose_json`. Not returned by
+          # `openai/whisper-large-v3-turbo`.
           duration: nil,
           # Segments of the transcribed text and their corresponding details. Returned by
-          # `distil-whisper/distil-large-v2` when `response_format` is `verbose_json`. Not
-          # returned by `openai/whisper-large-v3-turbo`.
+          # `distil-whisper/distil-large-v2` and the `deepgram/*` models when
+          # `response_format` is `verbose_json`; Deepgram segments also carry nested `words`
+          # and `speakers`. Not returned by `openai/whisper-large-v3-turbo`.
           segments: nil,
-          # Word-level timestamps and optional speaker labels. Only returned by
-          # `deepgram/nova-3` when word-level output is enabled via `model_config`.
+          # Word-level timestamps and optional speaker labels. Only returned by the
+          # `deepgram/*` models when word-level output is enabled via `model_config`.
           words: nil
         )
         end
@@ -111,7 +108,7 @@ module Telnyx
               duration: Float,
               segments:
                 T::Array[Telnyx::Models::AI::AudioTranscribeResponse::Segment],
-              words: T::Array[Telnyx::Models::AI::AudioTranscribeResponse::Word]
+              words: T::Array[Telnyx::AI::AudioTranscriptionResponseWord]
             }
           )
         end
@@ -143,100 +140,69 @@ module Telnyx
           sig { returns(String) }
           attr_accessor :text
 
+          # Speaker indices heard in this segment. Returned by the `deepgram/*` models when
+          # `diarize` is enabled via `model_config`.
+          sig { returns(T.nilable(T::Array[Integer])) }
+          attr_reader :speakers
+
+          sig { params(speakers: T::Array[Integer]).void }
+          attr_writer :speakers
+
+          # Word-level timing detail for this segment. Returned by the `deepgram/*` models
+          # when word-level output is enabled via `model_config`.
           sig do
-            params(id: Float, end_: Float, start: Float, text: String).returns(
-              T.attached_class
+            returns(
+              T.nilable(T::Array[Telnyx::AI::AudioTranscriptionResponseWord])
             )
           end
-          def self.new(
-            # Unique identifier of the segment.
-            id:,
-            # End time of the segment in seconds.
-            end_:,
-            # Start time of the segment in seconds.
-            start:,
-            # Text content of the segment.
-            text:
-          )
-          end
+          attr_reader :words
 
           sig do
-            override.returns(
-              { id: Float, end_: Float, start: Float, text: String }
-            )
-          end
-          def to_hash
+            params(
+              words:
+                T::Array[Telnyx::AI::AudioTranscriptionResponseWord::OrHash]
+            ).void
           end
-        end
+          attr_writer :words
 
-        class Word < Telnyx::Internal::Type::BaseModel
-          OrHash =
-            T.type_alias do
-              T.any(
-                Telnyx::Models::AI::AudioTranscribeResponse::Word,
-                Telnyx::Internal::AnyHash
-              )
-            end
-
-          # End time of the word in seconds.
-          sig { returns(Float) }
-          attr_accessor :end_
-
-          # Start time of the word in seconds.
-          sig { returns(Float) }
-          attr_accessor :start
-
-          # The transcribed word.
-          sig { returns(String) }
-          attr_accessor :word
-
-          # Confidence score for the word (0.0 to 1.0).
-          sig { returns(T.nilable(Float)) }
-          attr_reader :confidence
-
-          sig { params(confidence: Float).void }
-          attr_writer :confidence
-
-          # Speaker index. Only present when diarization is enabled via `model_config`.
-          sig { returns(T.nilable(Integer)) }
-          attr_reader :speaker
-
-          sig { params(speaker: Integer).void }
-          attr_writer :speaker
-
-          # Word-level timing detail. Only present when using `deepgram/nova-3` with
-          # `model_config` options that enable word timestamps.
           sig do
             params(
+              id: Float,
               end_: Float,
               start: Float,
-              word: String,
-              confidence: Float,
-              speaker: Integer
+              text: String,
+              speakers: T::Array[Integer],
+              words:
+                T::Array[Telnyx::AI::AudioTranscriptionResponseWord::OrHash]
             ).returns(T.attached_class)
           end
           def self.new(
-            # End time of the word in seconds.
+            # Unique identifier of the segment.
+            id:,
+            # End time of the segment in seconds.
             end_:,
-            # Start time of the word in seconds.
+            # Start time of the segment in seconds.
             start:,
-            # The transcribed word.
-            word:,
-            # Confidence score for the word (0.0 to 1.0).
-            confidence: nil,
-            # Speaker index. Only present when diarization is enabled via `model_config`.
-            speaker: nil
+            # Text content of the segment.
+            text:,
+            # Speaker indices heard in this segment. Returned by the `deepgram/*` models when
+            # `diarize` is enabled via `model_config`.
+            speakers: nil,
+            # Word-level timing detail for this segment. Returned by the `deepgram/*` models
+            # when word-level output is enabled via `model_config`.
+            words: nil
           )
           end
 
           sig do
             override.returns(
               {
+                id: Float,
                 end_: Float,
                 start: Float,
-                word: String,
-                confidence: Float,
-                speaker: Integer
+                text: String,
+                speakers: T::Array[Integer],
+                words: T::Array[Telnyx::AI::AudioTranscriptionResponseWord]
               }
             )
           end

data/rbi/telnyx/models/ai/audio_transcription_response_word.rbi

@@ -0,0 +1,108 @@
+# typed: strong
+
+module Telnyx
+  module Models
+    module AI
+      class AudioTranscriptionResponseWord < Telnyx::Internal::Type::BaseModel
+        OrHash =
+          T.type_alias do
+            T.any(
+              Telnyx::AI::AudioTranscriptionResponseWord,
+              Telnyx::Internal::AnyHash
+            )
+          end
+
+        # End time of the word in seconds.
+        sig { returns(Float) }
+        attr_accessor :end_
+
+        # Start time of the word in seconds.
+        sig { returns(Float) }
+        attr_accessor :start
+
+        # The transcribed word.
+        sig { returns(String) }
+        attr_accessor :word
+
+        # Confidence score for the word (0.0 to 1.0).
+        sig { returns(T.nilable(Float)) }
+        attr_reader :confidence
+
+        sig { params(confidence: Float).void }
+        attr_writer :confidence
+
+        # The transcribed word with punctuation and capitalisation applied. Only present
+        # when `punctuate` or `smart_format` is enabled via `model_config`.
+        sig { returns(T.nilable(String)) }
+        attr_reader :punctuated_word
+
+        sig { params(punctuated_word: String).void }
+        attr_writer :punctuated_word
+
+        # Speaker index. Only present when diarization is enabled via `model_config`.
+        sig { returns(T.nilable(Integer)) }
+        attr_reader :speaker
+
+        sig { params(speaker: Integer).void }
+        attr_writer :speaker
+
+        # Confidence score for the speaker assignment (0.0 to 1.0). Only present when
+        # diarization is enabled via `model_config`.
+        sig { returns(T.nilable(Float)) }
+        attr_reader :speaker_confidence
+
+        sig { params(speaker_confidence: Float).void }
+        attr_writer :speaker_confidence
+
+        # Word-level timing detail. Only present when using a `deepgram/*` model with
+        # `model_config` options that enable word timestamps.
+        sig do
+          params(
+            end_: Float,
+            start: Float,
+            word: String,
+            confidence: Float,
+            punctuated_word: String,
+            speaker: Integer,
+            speaker_confidence: Float
+          ).returns(T.attached_class)
+        end
+        def self.new(
+          # End time of the word in seconds.
+          end_:,
+          # Start time of the word in seconds.
+          start:,
+          # The transcribed word.
+          word:,
+          # Confidence score for the word (0.0 to 1.0).
+          confidence: nil,
+          # The transcribed word with punctuation and capitalisation applied. Only present
+          # when `punctuate` or `smart_format` is enabled via `model_config`.
+          punctuated_word: nil,
+          # Speaker index. Only present when diarization is enabled via `model_config`.
+          speaker: nil,
+          # Confidence score for the speaker assignment (0.0 to 1.0). Only present when
+          # diarization is enabled via `model_config`.
+          speaker_confidence: nil
+        )
+        end
+
+        sig do
+          override.returns(
+            {
+              end_: Float,
+              start: Float,
+              word: String,
+              confidence: Float,
+              punctuated_word: String,
+              speaker: Integer,
+              speaker_confidence: Float
+            }
+          )
+        end
+        def to_hash
+        end
+      end
+    end
+  end
+end