telnyx 5.171.0 → 5.174.0

data/lib/telnyx/models/ai/transcription_challenger.rb

@@ -0,0 +1,148 @@
+# frozen_string_literal: true
+
+module Telnyx
+  module Models
+    module AI
+      class TranscriptionChallenger < Telnyx::Internal::Type::BaseModel
+        # @!attribute model
+        #   The language booster's model. It must be the same kind of model as
+        #   `transcription.model`: both streaming (`deepgram/flux`, `deepgram/nova-3`,
+        #   `deepgram/nova-2`, `assemblyai/universal-3-6-pro` or its legacy aliases
+        #   `assemblyai/universal-3-5-pro` and `assemblyai/universal-streaming`,
+        #   `xai/grok-stt`, `soniox/stt-rt-v4`, `soniox/stt-rt-v5`, `humain/realtime`,
+        #   `reson8/turns`) or both non-streaming (`azure/fast`, `nvidia/parakeet-v3`,
+        #   `omi-health/omi-med-stt-v1`, `cohere/ar-stt`, `distil-whisper/distil-large-v2`,
+        #   `openai/whisper-large-v3-turbo`, `telnyx/basira`). It can be the same model as
+        #   `transcription.model` on a different `language`.
+        #
+        #   @return [Symbol, Telnyx::Models::AI::TranscriptionChallenger::Model]
+        required :model, enum: -> { Telnyx::AI::TranscriptionChallenger::Model }
+
+        # @!attribute language
+        #   The language this model transcribes. Omit it or set it to `null` to use the
+        #   language of `transcription.model`. The request is rejected when this model
+        #   doesn't support the language it would run. It is also rejected when it would run
+        #   the same model on the same language as `transcription.model`.
+        #
+        #   @return [String, nil]
+        optional :language, String, nil?: true
+
+        # @!attribute rule
+        #   How the assistant picks the transcript it uses. The models are compared on how
+        #   complete and confident their transcripts are, not on language, so the rules work
+        #   best when both models understand the callers' language.
+        #
+        #   - `best_turn` (default): both models transcribe the whole call. Each turn uses
+        #     the language booster's transcript only when it scores higher than the
+        #     transcript of `transcription.model` (clearly higher with non-streaming
+        #     models). With streaming models, `transcription.model` also decides when each
+        #     turn ends. Available for every pair.
+        #   - `best_engine`: both models transcribe the first turns, then the call continues
+        #     alone on the model whose transcripts scored higher. If neither clearly leads,
+        #     `transcription.model` continues. Streaming models only.
+        #   - `merge_words`: both models transcribe each utterance and their words are
+        #     merged, keeping Arabic and English spoken in the same sentence. Available only
+        #     for `telnyx/basira` with `cohere/ar-stt`, in either order. The pair runs on
+        #     the language that applies to `telnyx/basira` (its own, or that of
+        #     `transcription.model`), which must be Arabic (`ar` or an `ar-` locale),
+        #     `multi`, or `auto`.
+        #
+        #   @return [Symbol, Telnyx::Models::AI::TranscriptionChallenger::Rule, nil]
+        optional :rule, enum: -> { Telnyx::AI::TranscriptionChallenger::Rule }
+
+        # @!attribute settings
+        #   Settings for the language booster, with the same fields and limits as
+        #   `transcription.settings`. Fields that don't apply to this model's provider are
+        #   dropped, and the provider's defaults fill in the rest. Omit it or set it to
+        #   `null` to use the settings of `transcription.model` where they apply to this
+        #   model.
+        #
+        #   @return [Telnyx::Models::AI::TranscriptionSettingsConfig, nil]
+        optional :settings, -> { Telnyx::AI::TranscriptionSettingsConfig }, nil?: true
+
+        # @!method initialize(model:, language: nil, rule: nil, settings: nil)
+        #   Some parameter documentations has been truncated, see
+        #   {Telnyx::Models::AI::TranscriptionChallenger} for more details.
+        #
+        #   A second speech-to-text model that transcribes alongside `transcription.model`,
+        #   and the rule that decides which transcript the assistant uses.
+        #
+        #   @param model [Symbol, Telnyx::Models::AI::TranscriptionChallenger::Model] The language booster's model. It must be the same kind of model as `transcriptio
+        #
+        #   @param language [String, nil] The language this model transcribes. Omit it or set it to `null` to use the lang
+        #
+        #   @param rule [Symbol, Telnyx::Models::AI::TranscriptionChallenger::Rule] How the assistant picks the transcript it uses. The models are compared on how c
+        #
+        #   @param settings [Telnyx::Models::AI::TranscriptionSettingsConfig, nil] Settings for the language booster, with the same fields and limits as `transcrip
+
+        # The language booster's model. It must be the same kind of model as
+        # `transcription.model`: both streaming (`deepgram/flux`, `deepgram/nova-3`,
+        # `deepgram/nova-2`, `assemblyai/universal-3-6-pro` or its legacy aliases
+        # `assemblyai/universal-3-5-pro` and `assemblyai/universal-streaming`,
+        # `xai/grok-stt`, `soniox/stt-rt-v4`, `soniox/stt-rt-v5`, `humain/realtime`,
+        # `reson8/turns`) or both non-streaming (`azure/fast`, `nvidia/parakeet-v3`,
+        # `omi-health/omi-med-stt-v1`, `cohere/ar-stt`, `distil-whisper/distil-large-v2`,
+        # `openai/whisper-large-v3-turbo`, `telnyx/basira`). It can be the same model as
+        # `transcription.model` on a different `language`.
+        #
+        # @see Telnyx::Models::AI::TranscriptionChallenger#model
+        module Model
+          extend Telnyx::Internal::Type::Enum
+
+          DEEPGRAM_FLUX = :"deepgram/flux"
+          DEEPGRAM_NOVA_3 = :"deepgram/nova-3"
+          DEEPGRAM_NOVA_2 = :"deepgram/nova-2"
+          AZURE_FAST = :"azure/fast"
+          ASSEMBLYAI_UNIVERSAL_3_6_PRO = :"assemblyai/universal-3-6-pro"
+          ASSEMBLYAI_UNIVERSAL_3_5_PRO = :"assemblyai/universal-3-5-pro"
+          ASSEMBLYAI_UNIVERSAL_STREAMING = :"assemblyai/universal-streaming"
+          XAI_GROK_STT = :"xai/grok-stt"
+          SONIOX_STT_RT_V4 = :"soniox/stt-rt-v4"
+          SONIOX_STT_RT_V5 = :"soniox/stt-rt-v5"
+          NVIDIA_PARAKEET_V3 = :"nvidia/parakeet-v3"
+          OMI_HEALTH_OMI_MED_STT_V1 = :"omi-health/omi-med-stt-v1"
+          HUMAIN_REALTIME = :"humain/realtime"
+          RESON8_TURNS = :"reson8/turns"
+          COHERE_AR_STT = :"cohere/ar-stt"
+          TELNYX_BASIRA = :"telnyx/basira"
+          DISTIL_WHISPER_DISTIL_LARGE_V2 = :"distil-whisper/distil-large-v2"
+          OPENAI_WHISPER_LARGE_V3_TURBO = :"openai/whisper-large-v3-turbo"
+
+          # @!method self.values
+          #   @return [Array<Symbol>]
+        end
+
+        # How the assistant picks the transcript it uses. The models are compared on how
+        # complete and confident their transcripts are, not on language, so the rules work
+        # best when both models understand the callers' language.
+        #
+        # - `best_turn` (default): both models transcribe the whole call. Each turn uses
+        #   the language booster's transcript only when it scores higher than the
+        #   transcript of `transcription.model` (clearly higher with non-streaming
+        #   models). With streaming models, `transcription.model` also decides when each
+        #   turn ends. Available for every pair.
+        # - `best_engine`: both models transcribe the first turns, then the call continues
+        #   alone on the model whose transcripts scored higher. If neither clearly leads,
+        #   `transcription.model` continues. Streaming models only.
+        # - `merge_words`: both models transcribe each utterance and their words are
+        #   merged, keeping Arabic and English spoken in the same sentence. Available only
+        #   for `telnyx/basira` with `cohere/ar-stt`, in either order. The pair runs on
+        #   the language that applies to `telnyx/basira` (its own, or that of
+        #   `transcription.model`), which must be Arabic (`ar` or an `ar-` locale),
+        #   `multi`, or `auto`.
+        #
+        # @see Telnyx::Models::AI::TranscriptionChallenger#rule
+        module Rule
+          extend Telnyx::Internal::Type::Enum
+
+          BEST_TURN = :best_turn
+          BEST_ENGINE = :best_engine
+          MERGE_WORDS = :merge_words
+
+          # @!method self.values
+          #   @return [Array<Symbol>]
+        end
+      end
+    end
+  end
+end

data/lib/telnyx/models/ai/transcription_fallback_model.rb

@@ -0,0 +1,78 @@
+# frozen_string_literal: true
+
+module Telnyx
+  module Models
+    module AI
+      class TranscriptionFallbackModel < Telnyx::Internal::Type::BaseModel
+        # @!attribute model
+        #   The fallback model. It must be a streaming model other than
+        #   `transcription.model` and the other fallbacks: `deepgram/flux`,
+        #   `deepgram/nova-3`, `deepgram/nova-2`, `assemblyai/universal-3-6-pro` (or its
+        #   legacy aliases `assemblyai/universal-3-5-pro` and
+        #   `assemblyai/universal-streaming`), `xai/grok-stt`, `soniox/stt-rt-v4`,
+        #   `soniox/stt-rt-v5`, `humain/realtime`, or `reson8/turns`.
+        #
+        #   @return [Symbol, Telnyx::Models::AI::TranscriptionFallbackModel::Model]
+        required :model, enum: -> { Telnyx::AI::TranscriptionFallbackModel::Model }
+
+        # @!attribute language
+        #   The language the fallback transcribes. Omit it or set it to `null` to use the
+        #   language of `transcription.model`. The request is rejected when the fallback
+        #   model doesn't support the language it would run.
+        #
+        #   @return [String, nil]
+        optional :language, String, nil?: true
+
+        # @!attribute settings
+        #   Settings for the fallback, with the same fields and limits as
+        #   `transcription.settings`. Fields that don't apply to this model's provider are
+        #   dropped, and the provider's defaults fill in the rest. Omit it or set it to
+        #   `null` to use the settings of `transcription.model` where they apply to this
+        #   model.
+        #
+        #   @return [Telnyx::Models::AI::TranscriptionSettingsConfig, nil]
+        optional :settings, -> { Telnyx::AI::TranscriptionSettingsConfig }, nil?: true
+
+        # @!method initialize(model:, language: nil, settings: nil)
+        #   Some parameter documentations has been truncated, see
+        #   {Telnyx::Models::AI::TranscriptionFallbackModel} for more details.
+        #
+        #   A streaming speech-to-text model that takes over transcription when the model in
+        #   use fails.
+        #
+        #   @param model [Symbol, Telnyx::Models::AI::TranscriptionFallbackModel::Model] The fallback model. It must be a streaming model other than `transcription.model
+        #
+        #   @param language [String, nil] The language the fallback transcribes. Omit it or set it to `null` to use the la
+        #
+        #   @param settings [Telnyx::Models::AI::TranscriptionSettingsConfig, nil] Settings for the fallback, with the same fields and limits as `transcription.set
+
+        # The fallback model. It must be a streaming model other than
+        # `transcription.model` and the other fallbacks: `deepgram/flux`,
+        # `deepgram/nova-3`, `deepgram/nova-2`, `assemblyai/universal-3-6-pro` (or its
+        # legacy aliases `assemblyai/universal-3-5-pro` and
+        # `assemblyai/universal-streaming`), `xai/grok-stt`, `soniox/stt-rt-v4`,
+        # `soniox/stt-rt-v5`, `humain/realtime`, or `reson8/turns`.
+        #
+        # @see Telnyx::Models::AI::TranscriptionFallbackModel#model
+        module Model
+          extend Telnyx::Internal::Type::Enum
+
+          DEEPGRAM_FLUX = :"deepgram/flux"
+          DEEPGRAM_NOVA_3 = :"deepgram/nova-3"
+          DEEPGRAM_NOVA_2 = :"deepgram/nova-2"
+          ASSEMBLYAI_UNIVERSAL_3_6_PRO = :"assemblyai/universal-3-6-pro"
+          ASSEMBLYAI_UNIVERSAL_3_5_PRO = :"assemblyai/universal-3-5-pro"
+          ASSEMBLYAI_UNIVERSAL_STREAMING = :"assemblyai/universal-streaming"
+          XAI_GROK_STT = :"xai/grok-stt"
+          SONIOX_STT_RT_V4 = :"soniox/stt-rt-v4"
+          SONIOX_STT_RT_V5 = :"soniox/stt-rt-v5"
+          HUMAIN_REALTIME = :"humain/realtime"
+          RESON8_TURNS = :"reson8/turns"
+
+          # @!method self.values
+          #   @return [Array<Symbol>]
+        end
+      end
+    end
+  end
+end

data/lib/telnyx/models/ai/transcription_settings.rb

@@ -15,8 +15,8 @@ module Telnyx
         #   A second speech-to-text model that transcribes alongside `transcription.model`,
         #   and the rule that decides which transcript the assistant uses.
         #
-        #   @return [Telnyx::Models::AI::TranscriptionSettings::Challenger, nil]
-        optional :challenger, -> { Telnyx::AI::TranscriptionSettings::Challenger }, nil?: true
+        #   @return [Telnyx::Models::AI::TranscriptionChallenger, nil]
+        optional :challenger, -> { Telnyx::AI::TranscriptionChallenger }, nil?: true
 
         # @!attribute fallback_models
         #   Up to 3 streaming models that take over transcription, in this order, when the
@@ -28,9 +28,9 @@ module Telnyx
         #   Can't be combined with `challenger`, the language booster; to replace a stored
         #   language booster, send `challenger: null` in the same request.
         #
-        #   @return [Array<Telnyx::Models::AI::TranscriptionSettings::FallbackModel>, nil]
+        #   @return [Array<Telnyx::Models::AI::TranscriptionFallbackModel>, nil]
         optional :fallback_models,
-                 -> { Telnyx::Internal::Type::ArrayOf[Telnyx::AI::TranscriptionSettings::FallbackModel] },
+                 -> { Telnyx::Internal::Type::ArrayOf[Telnyx::AI::TranscriptionFallbackModel] },
                  nil?: true
 
         # @!attribute language
@@ -50,7 +50,12 @@ module Telnyx
         #   `nl`, `en`, `fr`, `fy`, `de`, `it`, `pl`, `pt`, `es`, and `sv` to fix the
         #   transcription language. For `cohere/ar-stt`, supported values are `ar` and `en`;
         #   unlike other models, this model does not auto-detect and defaults to `ar` when
-        #   `language` is omitted.
+        #   `language` is omitted. For `speechmatics/linden-1`, there is no auto-detect; set
+        #   a code from Speechmatics' Realtime language list (67 language codes, including
+        #   `es-bilingual-en` for Spanish and English bilingual, and the spelling and script
+        #   variants `en-US`, `en-GB`, `en-AU`, `cmn-Hans`, and `cmn-Hant`; `en` alone uses
+        #   the default spelling); unset defaults to `en`, while an explicit `auto`,
+        #   `auto_beta`, or `multi` is rejected with a 400.
         #
         #   @return [String, nil]
         optional :language, String
@@ -62,10 +67,17 @@ module Telnyx
         #   - `deepgram/flux` is optimized for turn-taking with multilingual language hints.
         #   - `deepgram/nova-3` is multilingual with automatic language detection.
         #   - `deepgram/nova-2` is Deepgram's previous-generation multilingual model.
+        #   - `speechmatics/standard` is a Speechmatics model self-hosted by Telnyx; it
+        #     supports a fixed set of language packs (including bilingual packs such as
+        #     `ar_en`) and defaults to `en`.
+        #   - `speechmatics/linden-1` (preview) is a vendor-hosted streaming model with no
+        #     auto-detect; set `language` from Speechmatics' Realtime language list
+        #     (defaults to `en`).
         #   - `azure/fast` is a multilingual Azure transcription model.
-        #   - `assemblyai/universal-3-5-pro` is a multilingual streaming model with
-        #     configurable turn detection. The legacy alias `assemblyai/universal-streaming`
-        #     is still accepted and resolves to the same model.
+        #   - `assemblyai/universal-3-6-pro` (AssemblyAI Universal-3.6 Pro) is a
+        #     multilingual streaming model with configurable turn detection. The legacy
+        #     aliases `assemblyai/universal-3-5-pro` and `assemblyai/universal-streaming`
+        #     are still accepted and run the same model.
         #   - `xai/grok-stt` is a multilingual Grok STT model.
         #   - `soniox/stt-rt-v4` and `soniox/stt-rt-v5` are multilingual streaming models
         #     with automatic language detection, configurable endpointing, term biasing
@@ -102,9 +114,9 @@ module Telnyx
         #
         #   @param api_key_ref [String] Integration secret identifier for the transcription provider API key. Currently
         #
-        #   @param challenger [Telnyx::Models::AI::TranscriptionSettings::Challenger, nil] A second speech-to-text model that transcribes alongside `transcription.model`,
+        #   @param challenger [Telnyx::Models::AI::TranscriptionChallenger, nil] A second speech-to-text model that transcribes alongside `transcription.model`,
         #
-        #   @param fallback_models [Array<Telnyx::Models::AI::TranscriptionSettings::FallbackModel>, nil] Up to 3 streaming models that take over transcription, in this order, when the m
+        #   @param fallback_models [Array<Telnyx::Models::AI::TranscriptionFallbackModel>, nil] Up to 3 streaming models that take over transcription, in this order, when the m
         #
         #   @param language [String] The language of the audio to be transcribed. If not set, or if set to `auto`, su
         #
@@ -114,225 +126,23 @@ module Telnyx
         #
         #   @param settings [Telnyx::Models::AI::TranscriptionSettingsConfig]
 
-        # @see Telnyx::Models::AI::TranscriptionSettings#challenger
-        class Challenger < Telnyx::Internal::Type::BaseModel
-          # @!attribute model
-          #   The language booster's model. It must be the same kind of model as
-          #   `transcription.model`: both streaming (`deepgram/flux`, `deepgram/nova-3`,
-          #   `deepgram/nova-2`, `assemblyai/universal-3-5-pro` or its legacy alias
-          #   `assemblyai/universal-streaming`, `xai/grok-stt`, `soniox/stt-rt-v4`,
-          #   `soniox/stt-rt-v5`, `humain/realtime`, `reson8/turns`) or both non-streaming
-          #   (`azure/fast`, `nvidia/parakeet-v3`, `omi-health/omi-med-stt-v1`,
-          #   `cohere/ar-stt`, `distil-whisper/distil-large-v2`,
-          #   `openai/whisper-large-v3-turbo`, `telnyx/basira`). It can be the same model as
-          #   `transcription.model` on a different `language`.
-          #
-          #   @return [Symbol, Telnyx::Models::AI::TranscriptionSettings::Challenger::Model]
-          required :model, enum: -> { Telnyx::AI::TranscriptionSettings::Challenger::Model }
-
-          # @!attribute language
-          #   The language this model transcribes. Omit it or set it to `null` to use the
-          #   language of `transcription.model`. The request is rejected when this model
-          #   doesn't support the language it would run. It is also rejected when it would run
-          #   the same model on the same language as `transcription.model`.
-          #
-          #   @return [String, nil]
-          optional :language, String, nil?: true
-
-          # @!attribute rule
-          #   How the assistant picks the transcript it uses. The models are compared on how
-          #   complete and confident their transcripts are, not on language, so the rules work
-          #   best when both models understand the callers' language.
-          #
-          #   - `best_turn` (default): both models transcribe the whole call. Each turn uses
-          #     the language booster's transcript only when it scores higher than the
-          #     transcript of `transcription.model` (clearly higher with non-streaming
-          #     models). With streaming models, `transcription.model` also decides when each
-          #     turn ends. Available for every pair.
-          #   - `best_engine`: both models transcribe the first turns, then the call continues
-          #     alone on the model whose transcripts scored higher. If neither clearly leads,
-          #     `transcription.model` continues. Streaming models only.
-          #   - `merge_words`: both models transcribe each utterance and their words are
-          #     merged, keeping Arabic and English spoken in the same sentence. Available only
-          #     for `telnyx/basira` with `cohere/ar-stt`, in either order. The pair runs on
-          #     the language that applies to `telnyx/basira` (its own, or that of
-          #     `transcription.model`), which must be Arabic (`ar` or an `ar-` locale),
-          #     `multi`, or `auto`.
-          #
-          #   @return [Symbol, Telnyx::Models::AI::TranscriptionSettings::Challenger::Rule, nil]
-          optional :rule, enum: -> { Telnyx::AI::TranscriptionSettings::Challenger::Rule }
-
-          # @!attribute settings
-          #   Settings for the language booster, with the same fields and limits as
-          #   `transcription.settings`. Fields that don't apply to this model's provider are
-          #   dropped, and the provider's defaults fill in the rest. Omit it or set it to
-          #   `null` to use the settings of `transcription.model` where they apply to this
-          #   model.
-          #
-          #   @return [Telnyx::Models::AI::TranscriptionSettingsConfig, nil]
-          optional :settings, -> { Telnyx::AI::TranscriptionSettingsConfig }, nil?: true
-
-          # @!method initialize(model:, language: nil, rule: nil, settings: nil)
-          #   Some parameter documentations has been truncated, see
-          #   {Telnyx::Models::AI::TranscriptionSettings::Challenger} for more details.
-          #
-          #   A second speech-to-text model that transcribes alongside `transcription.model`,
-          #   and the rule that decides which transcript the assistant uses.
-          #
-          #   @param model [Symbol, Telnyx::Models::AI::TranscriptionSettings::Challenger::Model] The language booster's model. It must be the same kind of model as `transcriptio
-          #
-          #   @param language [String, nil] The language this model transcribes. Omit it or set it to `null` to use the lang
-          #
-          #   @param rule [Symbol, Telnyx::Models::AI::TranscriptionSettings::Challenger::Rule] How the assistant picks the transcript it uses. The models are compared on how c
-          #
-          #   @param settings [Telnyx::Models::AI::TranscriptionSettingsConfig, nil] Settings for the language booster, with the same fields and limits as `transcrip
-
-          # The language booster's model. It must be the same kind of model as
-          # `transcription.model`: both streaming (`deepgram/flux`, `deepgram/nova-3`,
-          # `deepgram/nova-2`, `assemblyai/universal-3-5-pro` or its legacy alias
-          # `assemblyai/universal-streaming`, `xai/grok-stt`, `soniox/stt-rt-v4`,
-          # `soniox/stt-rt-v5`, `humain/realtime`, `reson8/turns`) or both non-streaming
-          # (`azure/fast`, `nvidia/parakeet-v3`, `omi-health/omi-med-stt-v1`,
-          # `cohere/ar-stt`, `distil-whisper/distil-large-v2`,
-          # `openai/whisper-large-v3-turbo`, `telnyx/basira`). It can be the same model as
-          # `transcription.model` on a different `language`.
-          #
-          # @see Telnyx::Models::AI::TranscriptionSettings::Challenger#model
-          module Model
-            extend Telnyx::Internal::Type::Enum
-
-            DEEPGRAM_FLUX = :"deepgram/flux"
-            DEEPGRAM_NOVA_3 = :"deepgram/nova-3"
-            DEEPGRAM_NOVA_2 = :"deepgram/nova-2"
-            AZURE_FAST = :"azure/fast"
-            ASSEMBLYAI_UNIVERSAL_3_5_PRO = :"assemblyai/universal-3-5-pro"
-            ASSEMBLYAI_UNIVERSAL_STREAMING = :"assemblyai/universal-streaming"
-            XAI_GROK_STT = :"xai/grok-stt"
-            SONIOX_STT_RT_V4 = :"soniox/stt-rt-v4"
-            SONIOX_STT_RT_V5 = :"soniox/stt-rt-v5"
-            NVIDIA_PARAKEET_V3 = :"nvidia/parakeet-v3"
-            OMI_HEALTH_OMI_MED_STT_V1 = :"omi-health/omi-med-stt-v1"
-            HUMAIN_REALTIME = :"humain/realtime"
-            RESON8_TURNS = :"reson8/turns"
-            COHERE_AR_STT = :"cohere/ar-stt"
-            TELNYX_BASIRA = :"telnyx/basira"
-            DISTIL_WHISPER_DISTIL_LARGE_V2 = :"distil-whisper/distil-large-v2"
-            OPENAI_WHISPER_LARGE_V3_TURBO = :"openai/whisper-large-v3-turbo"
-
-            # @!method self.values
-            #   @return [Array<Symbol>]
-          end
-
-          # How the assistant picks the transcript it uses. The models are compared on how
-          # complete and confident their transcripts are, not on language, so the rules work
-          # best when both models understand the callers' language.
-          #
-          # - `best_turn` (default): both models transcribe the whole call. Each turn uses
-          #   the language booster's transcript only when it scores higher than the
-          #   transcript of `transcription.model` (clearly higher with non-streaming
-          #   models). With streaming models, `transcription.model` also decides when each
-          #   turn ends. Available for every pair.
-          # - `best_engine`: both models transcribe the first turns, then the call continues
-          #   alone on the model whose transcripts scored higher. If neither clearly leads,
-          #   `transcription.model` continues. Streaming models only.
-          # - `merge_words`: both models transcribe each utterance and their words are
-          #   merged, keeping Arabic and English spoken in the same sentence. Available only
-          #   for `telnyx/basira` with `cohere/ar-stt`, in either order. The pair runs on
-          #   the language that applies to `telnyx/basira` (its own, or that of
-          #   `transcription.model`), which must be Arabic (`ar` or an `ar-` locale),
-          #   `multi`, or `auto`.
-          #
-          # @see Telnyx::Models::AI::TranscriptionSettings::Challenger#rule
-          module Rule
-            extend Telnyx::Internal::Type::Enum
-
-            BEST_TURN = :best_turn
-            BEST_ENGINE = :best_engine
-            MERGE_WORDS = :merge_words
-
-            # @!method self.values
-            #   @return [Array<Symbol>]
-          end
-        end
-
-        class FallbackModel < Telnyx::Internal::Type::BaseModel
-          # @!attribute model
-          #   The fallback model. It must be a streaming model other than
-          #   `transcription.model` and the other fallbacks: `deepgram/flux`,
-          #   `deepgram/nova-3`, `deepgram/nova-2`, `assemblyai/universal-3-5-pro` (or its
-          #   legacy alias `assemblyai/universal-streaming`), `xai/grok-stt`,
-          #   `soniox/stt-rt-v4`, `soniox/stt-rt-v5`, `humain/realtime`, or `reson8/turns`.
-          #
-          #   @return [Symbol, Telnyx::Models::AI::TranscriptionSettings::FallbackModel::Model]
-          required :model, enum: -> { Telnyx::AI::TranscriptionSettings::FallbackModel::Model }
-
-          # @!attribute language
-          #   The language the fallback transcribes. Omit it or set it to `null` to use the
-          #   language of `transcription.model`. The request is rejected when the fallback
-          #   model doesn't support the language it would run.
-          #
-          #   @return [String, nil]
-          optional :language, String, nil?: true
-
-          # @!attribute settings
-          #   Settings for the fallback, with the same fields and limits as
-          #   `transcription.settings`. Fields that don't apply to this model's provider are
-          #   dropped, and the provider's defaults fill in the rest. Omit it or set it to
-          #   `null` to use the settings of `transcription.model` where they apply to this
-          #   model.
-          #
-          #   @return [Telnyx::Models::AI::TranscriptionSettingsConfig, nil]
-          optional :settings, -> { Telnyx::AI::TranscriptionSettingsConfig }, nil?: true
-
-          # @!method initialize(model:, language: nil, settings: nil)
-          #   Some parameter documentations has been truncated, see
-          #   {Telnyx::Models::AI::TranscriptionSettings::FallbackModel} for more details.
-          #
-          #   A streaming speech-to-text model that takes over transcription when the model in
-          #   use fails.
-          #
-          #   @param model [Symbol, Telnyx::Models::AI::TranscriptionSettings::FallbackModel::Model] The fallback model. It must be a streaming model other than `transcription.model
-          #
-          #   @param language [String, nil] The language the fallback transcribes. Omit it or set it to `null` to use the la
-          #
-          #   @param settings [Telnyx::Models::AI::TranscriptionSettingsConfig, nil] Settings for the fallback, with the same fields and limits as `transcription.set
-
-          # The fallback model. It must be a streaming model other than
-          # `transcription.model` and the other fallbacks: `deepgram/flux`,
-          # `deepgram/nova-3`, `deepgram/nova-2`, `assemblyai/universal-3-5-pro` (or its
-          # legacy alias `assemblyai/universal-streaming`), `xai/grok-stt`,
-          # `soniox/stt-rt-v4`, `soniox/stt-rt-v5`, `humain/realtime`, or `reson8/turns`.
-          #
-          # @see Telnyx::Models::AI::TranscriptionSettings::FallbackModel#model
-          module Model
-            extend Telnyx::Internal::Type::Enum
-
-            DEEPGRAM_FLUX = :"deepgram/flux"
-            DEEPGRAM_NOVA_3 = :"deepgram/nova-3"
-            DEEPGRAM_NOVA_2 = :"deepgram/nova-2"
-            ASSEMBLYAI_UNIVERSAL_3_5_PRO = :"assemblyai/universal-3-5-pro"
-            ASSEMBLYAI_UNIVERSAL_STREAMING = :"assemblyai/universal-streaming"
-            XAI_GROK_STT = :"xai/grok-stt"
-            SONIOX_STT_RT_V4 = :"soniox/stt-rt-v4"
-            SONIOX_STT_RT_V5 = :"soniox/stt-rt-v5"
-            HUMAIN_REALTIME = :"humain/realtime"
-            RESON8_TURNS = :"reson8/turns"
-
-            # @!method self.values
-            #   @return [Array<Symbol>]
-          end
-        end
-
         # The speech to text model to be used by the voice assistant. All Deepgram models
         # are run on-premise.
         #
         # - `deepgram/flux` is optimized for turn-taking with multilingual language hints.
         # - `deepgram/nova-3` is multilingual with automatic language detection.
         # - `deepgram/nova-2` is Deepgram's previous-generation multilingual model.
+        # - `speechmatics/standard` is a Speechmatics model self-hosted by Telnyx; it
+        #   supports a fixed set of language packs (including bilingual packs such as
+        #   `ar_en`) and defaults to `en`.
+        # - `speechmatics/linden-1` (preview) is a vendor-hosted streaming model with no
+        #   auto-detect; set `language` from Speechmatics' Realtime language list
+        #   (defaults to `en`).
         # - `azure/fast` is a multilingual Azure transcription model.
-        # - `assemblyai/universal-3-5-pro` is a multilingual streaming model with
-        #   configurable turn detection. The legacy alias `assemblyai/universal-streaming`
-        #   is still accepted and resolves to the same model.
+        # - `assemblyai/universal-3-6-pro` (AssemblyAI Universal-3.6 Pro) is a
+        #   multilingual streaming model with configurable turn detection. The legacy
+        #   aliases `assemblyai/universal-3-5-pro` and `assemblyai/universal-streaming`
+        #   are still accepted and run the same model.
         # - `xai/grok-stt` is a multilingual Grok STT model.
         # - `soniox/stt-rt-v4` and `soniox/stt-rt-v5` are multilingual streaming models
         #   with automatic language detection, configurable endpointing, term biasing
@@ -355,7 +165,10 @@ module Telnyx
           DEEPGRAM_FLUX = :"deepgram/flux"
           DEEPGRAM_NOVA_3 = :"deepgram/nova-3"
           DEEPGRAM_NOVA_2 = :"deepgram/nova-2"
+          SPEECHMATICS_STANDARD = :"speechmatics/standard"
+          SPEECHMATICS_LINDEN_1 = :"speechmatics/linden-1"
           AZURE_FAST = :"azure/fast"
+          ASSEMBLYAI_UNIVERSAL_3_6_PRO = :"assemblyai/universal-3-6-pro"
           ASSEMBLYAI_UNIVERSAL_3_5_PRO = :"assemblyai/universal-3-5-pro"
           ASSEMBLYAI_UNIVERSAL_STREAMING = :"assemblyai/universal-streaming"
           XAI_GROK_STT = :"xai/grok-stt"