voice-router-dev 0.8.9 → 0.9.1

package/dist/index.mjs

@@ -754,60 +754,60 @@ var SonioxLanguage = {
 // src/generated/soniox/models.ts
 var SonioxModels = [
   { id: "stt-rt-v4", name: "Speech-to-Text Real-time v4", mode: "real_time" },
-  { id: "stt-rt-v3", name: "Speech-to-Text Real-time v3", mode: "real_time" },
   { id: "stt-async-v4", name: "Speech-to-Text Async v4", mode: "async" },
-  { id: "stt-async-v3", name: "Speech-to-Text Async v3", mode: "async" },
-  { id: "stt-rt-preview", name: "Speech-to-Text Real-time Preview", mode: "real_time", aliasOf: "stt-rt-v3" },
-  { id: "stt-async-preview", name: "Speech-to-Text Async Preview", mode: "async", aliasOf: "stt-async-v3" },
-  { id: "stt-rt-v3-preview", name: "Speech-to-Text Real-time v3 Preview", mode: "real_time", aliasOf: "stt-rt-v3" },
-  { id: "stt-rt-preview-v2", name: "Speech-to-Text Real-time Preview v2", mode: "real_time", aliasOf: "stt-rt-v3" },
-  { id: "stt-async-preview-v1", name: "Speech-to-Text Async Preview v1", mode: "async", aliasOf: "stt-async-v3" }
+  { id: "stt-rt-preview", name: "Speech-to-Text Real-time Preview", mode: "real_time", aliasOf: "stt-rt-v4" },
+  { id: "stt-async-preview", name: "Speech-to-Text Async Preview", mode: "async", aliasOf: "stt-async-v4" },
+  { id: "stt-rt-v3-preview", name: "Speech-to-Text Real-time v3 Preview", mode: "real_time", aliasOf: "stt-rt-v4" },
+  { id: "stt-rt-preview-v2", name: "Speech-to-Text Real-time Preview v2", mode: "real_time", aliasOf: "stt-rt-v4" },
+  { id: "stt-async-preview-v1", name: "Speech-to-Text Async Preview v1", mode: "async", aliasOf: "stt-async-v4" },
+  { id: "stt-rt-v3", name: "Speech-to-Text Real-time v3", mode: "real_time", aliasOf: "stt-rt-v4" },
+  { id: "stt-async-v3", name: "Speech-to-Text Async v3", mode: "async", aliasOf: "stt-async-v4" }
 ];
 var SonioxModelCodes = [
   "stt-rt-v4",
-  "stt-rt-v3",
   "stt-async-v4",
-  "stt-async-v3",
   "stt-rt-preview",
   "stt-async-preview",
   "stt-rt-v3-preview",
   "stt-rt-preview-v2",
-  "stt-async-preview-v1"
+  "stt-async-preview-v1",
+  "stt-rt-v3",
+  "stt-async-v3"
 ];
 var SonioxModelLabels = {
   "stt-rt-v4": "Speech-to-Text Real-time v4",
-  "stt-rt-v3": "Speech-to-Text Real-time v3",
   "stt-async-v4": "Speech-to-Text Async v4",
-  "stt-async-v3": "Speech-to-Text Async v3",
   "stt-rt-preview": "Speech-to-Text Real-time Preview",
   "stt-async-preview": "Speech-to-Text Async Preview",
   "stt-rt-v3-preview": "Speech-to-Text Real-time v3 Preview",
   "stt-rt-preview-v2": "Speech-to-Text Real-time Preview v2",
-  "stt-async-preview-v1": "Speech-to-Text Async Preview v1"
+  "stt-async-preview-v1": "Speech-to-Text Async Preview v1",
+  "stt-rt-v3": "Speech-to-Text Real-time v3",
+  "stt-async-v3": "Speech-to-Text Async v3"
 };
 var SonioxModel = {
   stt_rt_v4: "stt-rt-v4",
-  stt_rt_v3: "stt-rt-v3",
   stt_async_v4: "stt-async-v4",
-  stt_async_v3: "stt-async-v3",
   stt_rt_preview: "stt-rt-preview",
   stt_async_preview: "stt-async-preview",
   stt_rt_v3_preview: "stt-rt-v3-preview",
   stt_rt_preview_v2: "stt-rt-preview-v2",
-  stt_async_preview_v1: "stt-async-preview-v1"
+  stt_async_preview_v1: "stt-async-preview-v1",
+  stt_rt_v3: "stt-rt-v3",
+  stt_async_v3: "stt-async-v3"
 };
 var SonioxRealtimeModel = {
   stt_rt_v4: "stt-rt-v4",
-  stt_rt_v3: "stt-rt-v3",
   stt_rt_preview: "stt-rt-preview",
   stt_rt_v3_preview: "stt-rt-v3-preview",
-  stt_rt_preview_v2: "stt-rt-preview-v2"
+  stt_rt_preview_v2: "stt-rt-preview-v2",
+  stt_rt_v3: "stt-rt-v3"
 };
 var SonioxAsyncModel = {
   stt_async_v4: "stt-async-v4",
-  stt_async_v3: "stt-async-v3",
   stt_async_preview: "stt-async-preview",
-  stt_async_preview_v1: "stt-async-preview-v1"
+  stt_async_preview_v1: "stt-async-preview-v1",
+  stt_async_v3: "stt-async-v3"
 };
 
 // src/generated/speechmatics/languages.ts
@@ -3782,17 +3782,17 @@ var SummaryTypesEnum = {
   concise: "concise"
 };
 
-// src/generated/gladia/schema/transcriptMessageType.ts
-var TranscriptMessageType = {
-  transcript: "transcript"
-};
-
 // src/generated/gladia/schema/transcriptionControllerListV2KindItem.ts
 var TranscriptionControllerListV2KindItem = {
   "pre-recorded": "pre-recorded",
   live: "live"
 };
 
+// src/generated/gladia/schema/transcriptMessageType.ts
+var TranscriptMessageType = {
+  transcript: "transcript"
+};
+
 // src/generated/gladia/schema/translationMessageType.ts
 var TranslationMessageType = {
   translation: "translation"
@@ -4064,7 +4064,7 @@ var WebhookTranscriptionSuccessPayloadEvent = {
 
 // src/generated/gladia/api/gladiaControlAPI.ts
 var preRecordedControllerInitPreRecordedJobV2 = (initTranscriptionRequest, options) => {
-  return axios.post(`/v2/pre-recorded`, initTranscriptionRequest, options);
+  return axios.post("/v2/pre-recorded", initTranscriptionRequest, options);
 };
 var preRecordedControllerGetPreRecordedJobV2 = (id, options) => {
   return axios.get(`/v2/pre-recorded/${id}`, options);
@@ -4079,13 +4079,13 @@ var preRecordedControllerGetAudioV2 = (id, options) => {
   });
 };
 var transcriptionControllerListV2 = (params, options) => {
-  return axios.get(`/v2/transcription`, {
+  return axios.get("/v2/transcription", {
     ...options,
     params: { ...params, ...options?.params }
   });
 };
 var streamingControllerInitStreamingSessionV2 = (streamingRequest, params, options) => {
-  return axios.post(`/v2/live`, streamingRequest, {
+  return axios.post("/v2/live", streamingRequest, {
     ...options,
     params: { ...params, ...options?.params }
   });
@@ -4351,7 +4351,6 @@ var GladiaAdapter = class extends BaseAdapter {
         sentiment: result?.sentiment_analysis || void 0,
         audioToLlm: result?.audio_to_llm || void 0,
         chapters: result?.chapterization || void 0,
-        speakerReidentification: result?.speaker_reidentification || void 0,
         structuredData: result?.structured_data_extraction || void 0,
         customMetadata: response.custom_metadata || void 0
       },
@@ -5413,17 +5412,17 @@ var PiiPolicy = {
   zodiac_sign: "zodiac_sign"
 };
 
+// src/generated/assemblyai/schema/redactedAudioStatus.ts
+var RedactedAudioStatus = {
+  redacted_audio_ready: "redacted_audio_ready"
+};
+
 // src/generated/assemblyai/schema/redactPiiAudioQuality.ts
 var RedactPiiAudioQuality = {
   mp3: "mp3",
   wav: "wav"
 };
 
-// src/generated/assemblyai/schema/redactedAudioStatus.ts
-var RedactedAudioStatus = {
-  redacted_audio_ready: "redacted_audio_ready"
-};
-
 // src/generated/assemblyai/schema/sentiment.ts
 var Sentiment = {
   POSITIVE: "POSITIVE",
@@ -5487,10 +5486,10 @@ var TranscriptRemoveAudioTags = {
 
 // src/generated/assemblyai/api/assemblyAIAPI.ts
 var createTranscript = (transcriptParams, options) => {
-  return axios2.post(`/v2/transcript`, transcriptParams, options);
+  return axios2.post("/v2/transcript", transcriptParams, options);
 };
 var listTranscripts = (params, options) => {
-  return axios2.get(`/v2/transcript`, {
+  return axios2.get("/v2/transcript", {
     ...options,
     params: { ...params, ...options?.params }
   });
@@ -5836,23 +5835,22 @@ var AssemblyAIAdapter = class extends BaseAdapter {
         "AssemblyAI adapter currently only supports URL-based audio input. Use audio.type='url'"
       );
     }
-    const aaiOpts = { ...options?.assemblyai };
-    if ("speech_model" in aaiOpts && aaiOpts.speech_model != null) {
-      if (!aaiOpts.speech_models) {
-        aaiOpts.speech_models = [aaiOpts.speech_model];
-      }
-      delete aaiOpts.speech_model;
+    const passthrough = options?.assemblyai;
+    let speechModels;
+    if (passthrough?.speech_model != null && !passthrough.speech_models) {
+      speechModels = [passthrough.speech_model];
+    } else if (passthrough?.speech_models) {
+      speechModels = passthrough.speech_models;
     }
+    const { speech_model: _deprecated, ...typedOpts } = passthrough ?? {};
     const request = {
-      ...aaiOpts,
+      ...typedOpts,
       audio_url: audioUrl,
       // speech_models is required — default to universal-3-pro
-      speech_models: aaiOpts.speech_models ?? [
-        "universal-3-pro"
-      ],
+      speech_models: speechModels ?? ["universal-3-pro"],
       // Enable punctuation and formatting by default
-      punctuate: aaiOpts.punctuate ?? true,
-      format_text: aaiOpts.format_text ?? true
+      punctuate: typedOpts.punctuate ?? true,
+      format_text: typedOpts.format_text ?? true
     };
     if (options) {
       if (options.model) {
@@ -6567,8 +6565,10 @@ var DeepgramAdapter = class extends BaseAdapter {
   /**
    * Submit audio for transcription
    *
-   * Sends audio to Deepgram API for transcription. Deepgram processes
-   * synchronously and returns results immediately (no polling required).
+   * Sends audio to Deepgram API for transcription. Deepgram normally processes
+   * synchronously and returns results immediately. When `webhookUrl` is set,
+   * Deepgram can instead return an async callback acknowledgment containing a
+   * request ID.
    *
    * @param audio - Audio input (URL or file buffer)
    * @param options - Transcription options
@@ -6619,17 +6619,59 @@ var DeepgramAdapter = class extends BaseAdapter {
           { params }
         ).then((res) => res.data);
       } else if (audio.type === "file") {
-        response = await this.client.post("/listen", audio.file, {
-          params,
-          headers: {
-            "Content-Type": "audio/*"
+        response = await this.client.post(
+          "/listen",
+          audio.file,
+          {
+            params,
+            headers: {
+              "Content-Type": "audio/*"
+            }
           }
-        }).then((res) => res.data);
+        ).then((res) => res.data);
       } else {
         throw new Error(
           "Deepgram adapter does not support stream type for pre-recorded transcription. Use transcribeStream() for real-time streaming."
         );
       }
+      if (options?.webhookUrl) {
+        const requestId = ("request_id" in response ? response.request_id : void 0) || ("metadata" in response ? response.metadata?.request_id : void 0);
+        if (!requestId) {
+          return {
+            success: false,
+            provider: this.name,
+            error: {
+              code: "MISSING_REQUEST_ID",
+              message: "Deepgram callback mode did not return a request ID"
+            },
+            raw: response
+          };
+        }
+        return {
+          success: true,
+          provider: this.name,
+          data: {
+            id: requestId,
+            text: "",
+            status: "queued"
+          },
+          tracking: {
+            requestId
+          },
+          raw: response
+        };
+      }
+      if (!("results" in response) || !("metadata" in response)) {
+        return {
+          success: false,
+          provider: this.name,
+          error: {
+            code: "INVALID_RESPONSE",
+            message: "Deepgram did not return a synchronous transcription payload"
+          },
+          raw: response
+        };
+      }
       return this.normalizeResponse(response);
     } catch (error) {
       return this.createErrorResponse(error);
@@ -7290,7 +7332,8 @@ var DeepgramAdapter = class extends BaseAdapter {
         break;
       }
       case "Metadata": {
-        callbacks?.onMetadata?.(message);
+        const { type: _, ...metadata } = message;
+        callbacks?.onMetadata?.(metadata);
         break;
       }
       case "Error": {
@@ -7627,13 +7670,13 @@ var TextNormalizationKind = {
 
 // src/generated/azure/api/speechServicesAPIVersion32.ts
 var transcriptionsList = (params, options) => {
-  return axios4.get(`/transcriptions`, {
+  return axios4.get("/transcriptions", {
     ...options,
     params: { ...params, ...options?.params }
   });
 };
 var transcriptionsCreate = (transcription, options) => {
-  return axios4.post(`/transcriptions`, transcription, options);
+  return axios4.post("/transcriptions", transcription, options);
 };
 var transcriptionsGet = (id, options) => {
   return axios4.get(`/transcriptions/${id}`, options);
@@ -7648,13 +7691,13 @@ var transcriptionsListFiles = (id, params, options) => {
   });
 };
 var webHooksList = (params, options) => {
-  return axios4.get(`/webhooks`, {
+  return axios4.get("/webhooks", {
     ...options,
     params: { ...params, ...options?.params }
   });
 };
 var webHooksCreate = (webHook, options) => {
-  return axios4.post(`/webhooks`, webHook, options);
+  return axios4.post("/webhooks", webHook, options);
 };
 var webHooksDelete = (id, options) => {
   return axios4.delete(`/webhooks/${id}`, options);
@@ -7726,10 +7769,7 @@ var AzureSTTAdapter = class extends BaseAdapter {
         contentUrls: [audio.url],
         properties: this.buildTranscriptionProperties(options)
       };
-      const response = await transcriptionsCreate(
-        transcriptionRequest,
-        this.getAxiosConfig()
-      );
+      const response = await transcriptionsCreate(transcriptionRequest, this.getAxiosConfig());
       const transcription = response.data;
       const transcriptId = transcription.self?.split("/").pop() || "";
       return await this.pollForCompletion(transcriptId);
@@ -7782,7 +7822,7 @@ var AzureSTTAdapter = class extends BaseAdapter {
         this.getAxiosConfig()
       );
       const files = filesResponse.data?.values || [];
-      const resultFile = files.find((file) => file.kind === "Transcription");
+      const resultFile = files.find((file) => file.kind === FileKind.Transcription);
       if (!resultFile?.links?.contentUrl) {
         return {
           success: false,
@@ -8006,15 +8046,20 @@ var AzureSTTAdapter = class extends BaseAdapter {
     return properties;
   }
   /**
-   * Normalize Azure status to unified status
+   * Normalize Azure status to unified status using generated AzureStatus constants
    */
   normalizeStatus(status) {
-    const statusStr = status?.toString().toLowerCase() || "";
-    if (statusStr.includes("succeeded")) return "completed";
-    if (statusStr.includes("running")) return "processing";
-    if (statusStr.includes("notstarted")) return "queued";
-    if (statusStr.includes("failed")) return "error";
-    return "queued";
+    switch (status) {
+      case Status.Succeeded:
+        return "completed";
+      case Status.Running:
+        return "processing";
+      case Status.Failed:
+        return "error";
+      case Status.NotStarted:
+      default:
+        return "queued";
+    }
   }
   /**
    * Normalize Azure transcription response to unified format
@@ -8134,30 +8179,30 @@ function getAzureOpenAIRealtimeUrl(endpoint, deployment, apiVersion = "2024-10-0
 import axios6 from "axios";
 var createTranscription = (createTranscriptionRequest, options) => {
   const formData = new FormData();
-  formData.append(`file`, createTranscriptionRequest.file);
-  formData.append(`model`, createTranscriptionRequest.model);
+  formData.append("file", createTranscriptionRequest.file);
+  formData.append("model", createTranscriptionRequest.model);
   if (createTranscriptionRequest.language !== void 0) {
-    formData.append(`language`, createTranscriptionRequest.language);
+    formData.append("language", createTranscriptionRequest.language);
   }
   if (createTranscriptionRequest.prompt !== void 0) {
-    formData.append(`prompt`, createTranscriptionRequest.prompt);
+    formData.append("prompt", createTranscriptionRequest.prompt);
   }
   if (createTranscriptionRequest.response_format !== void 0) {
-    formData.append(`response_format`, createTranscriptionRequest.response_format);
+    formData.append("response_format", createTranscriptionRequest.response_format);
   }
   if (createTranscriptionRequest.temperature !== void 0) {
-    formData.append(`temperature`, createTranscriptionRequest.temperature.toString());
+    formData.append("temperature", createTranscriptionRequest.temperature.toString());
   }
   if (createTranscriptionRequest.include !== void 0) {
-    createTranscriptionRequest.include.forEach((value) => formData.append(`include`, value));
+    createTranscriptionRequest.include.forEach((value) => formData.append("include", value));
   }
   if (createTranscriptionRequest.timestamp_granularities !== void 0) {
     createTranscriptionRequest.timestamp_granularities.forEach(
-      (value) => formData.append(`timestamp_granularities`, value)
+      (value) => formData.append("timestamp_granularities", value)
     );
   }
   if (createTranscriptionRequest.stream !== void 0 && createTranscriptionRequest.stream !== null) {
-    formData.append(`stream`, createTranscriptionRequest.stream.toString());
+    formData.append("stream", createTranscriptionRequest.stream.toString());
   }
   if (createTranscriptionRequest.chunking_strategy !== void 0 && createTranscriptionRequest.chunking_strategy !== null) {
     formData.append(
@@ -8167,15 +8212,15 @@ var createTranscription = (createTranscriptionRequest, options) => {
   }
   if (createTranscriptionRequest.known_speaker_names !== void 0) {
     createTranscriptionRequest.known_speaker_names.forEach(
-      (value) => formData.append(`known_speaker_names`, value)
+      (value) => formData.append("known_speaker_names", value)
     );
   }
   if (createTranscriptionRequest.known_speaker_references !== void 0) {
     createTranscriptionRequest.known_speaker_references.forEach(
-      (value) => formData.append(`known_speaker_references`, value)
+      (value) => formData.append("known_speaker_references", value)
     );
   }
-  return axios6.post(`/audio/transcriptions`, formData, options);
+  return axios6.post("/audio/transcriptions", formData, options);
 };
 
 // src/generated/openai/schema/createTranscriptionRequestTimestampGranularitiesItem.ts
@@ -8264,7 +8309,6 @@ var OpenAIWhisperAdapter = class extends BaseAdapter {
       const request = {
         ...options?.openai,
         file: audioData,
-        // Generated type expects Blob
         model
       };
       if (options?.language) {
@@ -8691,7 +8735,6 @@ function createOpenAIWhisperAdapter(config) {
 
 // src/adapters/speechmatics-adapter.ts
 import axios8 from "axios";
-import WebSocket6 from "ws";
 
 // src/generated/speechmatics/schema/notificationConfigContentsItem.ts
 var NotificationConfigContentsItem = {
@@ -8741,7 +8784,8 @@ var SpeechmaticsAdapter = class extends BaseAdapter {
     super(...arguments);
     this.name = "speechmatics";
     this.capabilities = {
-      streaming: true,
+      streaming: false,
+      // Batch only (streaming available via separate WebSocket API)
       diarization: true,
       wordTimestamps: true,
       languageDetection: false,
@@ -8876,16 +8920,13 @@ var SpeechmaticsAdapter = class extends BaseAdapter {
         jobConfig.fetch_data = {
           url: audio.url
         };
-        const formData = new FormData();
-        formData.append("config", JSON.stringify(jobConfig));
-        requestBody = formData;
-        headers = { "Content-Type": "multipart/form-data" };
+        requestBody = { config: JSON.stringify(jobConfig) };
+        headers = { "Content-Type": "application/json" };
       } else if (audio.type === "file") {
-        const formData = new FormData();
-        formData.append("config", JSON.stringify(jobConfig));
-        const audioBlob = audio.file instanceof Blob ? audio.file : new Blob([audio.file], { type: audio.mimeType || "audio/wav" });
-        formData.append("data_file", audioBlob, audio.filename || "audio.wav");
-        requestBody = formData;
+        requestBody = {
+          config: JSON.stringify(jobConfig),
+          data_file: audio.file
+        };
         headers = { "Content-Type": "multipart/form-data" };
       } else {
         return {
@@ -8990,389 +9031,6 @@ var SpeechmaticsAdapter = class extends BaseAdapter {
       throw error;
     }
   }
-  /**
-   * Build WebSocket URL for real-time streaming
-   *
-   * Note: Real-time API uses a different host from the batch API:
-   * - Batch: {region}.asr.api.speechmatics.com
-   * - Real-time: {region}.rt.speechmatics.com
-   *
-   * @param region - Regional endpoint identifier
-   * @returns WebSocket URL for real-time API
-   */
-  getRegionalWsUrl(region) {
-    if (this.config?.wsBaseUrl) {
-      return this.config.wsBaseUrl;
-    }
-    const rtRegionMap = {
-      eu1: "eu",
-      eu2: "eu",
-      us1: "us",
-      us2: "us",
-      au1: "eu"
-      // No AU RT endpoint — fall back to EU
-    };
-    const rtPrefix = rtRegionMap[region || ""] || "eu";
-    return `wss://${rtPrefix}.rt.speechmatics.com/v2`;
-  }
-  /**
-   * Stream audio for real-time transcription via WebSocket
-   *
-   * Connects to Speechmatics' real-time API and sends audio chunks
-   * for transcription with results returned via callbacks.
-   *
-   * @param options - Streaming configuration options
-   * @param callbacks - Event callbacks for transcription results
-   * @returns Promise that resolves with a StreamingSession
-   *
-   * @example Basic streaming
-   * ```typescript
-   * const session = await adapter.transcribeStream({
-   *   language: 'en',
-   *   speechmaticsStreaming: {
-   *     enablePartials: true,
-   *     operatingPoint: 'enhanced'
-   *   }
-   * }, {
-   *   onTranscript: (event) => console.log(event.text),
-   *   onUtterance: (utt) => console.log(`[${utt.speaker}]: ${utt.text}`),
-   *   onError: (error) => console.error(error)
-   * });
-   *
-   * await session.sendAudio({ data: audioBuffer });
-   * await session.close();
-   * ```
-   */
-  async transcribeStream(options, callbacks) {
-    this.validateConfig();
-    const smOpts = options?.speechmaticsStreaming || {};
-    const region = smOpts.region || this.config?.region;
-    const wsUrl = this.getRegionalWsUrl(region);
-    const ws = new WebSocket6(wsUrl, {
-      headers: {
-        Authorization: `Bearer ${this.config.apiKey}`
-      }
-    });
-    let sessionStatus = "connecting";
-    const sessionId = `speechmatics-${Date.now()}-${Math.random().toString(36).substring(7)}`;
-    let seqNo = 0;
-    let utteranceResults = [];
-    const sessionReady = new Promise((resolve, reject) => {
-      const timeout = setTimeout(() => {
-        reject(new Error("WebSocket connection timeout"));
-      }, 1e4);
-      let wsOpen = false;
-      ws.once("error", (error) => {
-        clearTimeout(timeout);
-        reject(error);
-      });
-      ws.once("open", () => {
-        wsOpen = true;
-        const encoding = smOpts.encoding || options?.encoding || "pcm_s16le";
-        const sampleRate = smOpts.sampleRate || options?.sampleRate || 16e3;
-        const startMsg = {
-          message: "StartRecognition",
-          audio_format: {
-            type: "raw",
-            encoding,
-            sample_rate: sampleRate
-          },
-          transcription_config: {
-            language: smOpts.language || options?.language || "en",
-            enable_partials: smOpts.enablePartials ?? options?.interimResults ?? true
-          }
-        };
-        const txConfig = startMsg.transcription_config;
-        if (smOpts.domain) txConfig.domain = smOpts.domain;
-        if (smOpts.operatingPoint) txConfig.operating_point = smOpts.operatingPoint;
-        if (smOpts.maxDelay !== void 0) txConfig.max_delay = smOpts.maxDelay;
-        if (smOpts.maxDelayMode) txConfig.max_delay_mode = smOpts.maxDelayMode;
-        if (smOpts.enableEntities !== void 0) txConfig.enable_entities = smOpts.enableEntities;
-        if (smOpts.diarization === "speaker" || options?.diarization) {
-          txConfig.diarization = "speaker";
-          if (smOpts.maxSpeakers) {
-            txConfig.speaker_diarization_config = {
-              max_speakers: smOpts.maxSpeakers
-            };
-          } else if (options?.speakersExpected) {
-            txConfig.speaker_diarization_config = {
-              max_speakers: options.speakersExpected
-            };
-          }
-        }
-        if (smOpts.additionalVocab && smOpts.additionalVocab.length > 0) {
-          txConfig.additional_vocab = smOpts.additionalVocab.map((word) => ({
-            content: word
-          }));
-        } else if (options?.customVocabulary && options.customVocabulary.length > 0) {
-          txConfig.additional_vocab = options.customVocabulary.map((word) => ({
-            content: word
-          }));
-        }
-        if (smOpts.conversationConfig) {
-          txConfig.conversation_config = {
-            end_of_utterance_silence_trigger: smOpts.conversationConfig.endOfUtteranceSilenceTrigger
-          };
-        }
-        const startPayload = JSON.stringify(startMsg);
-        if (callbacks?.onRawMessage) {
-          callbacks.onRawMessage({
-            provider: "speechmatics",
-            direction: "outgoing",
-            timestamp: Date.now(),
-            payload: startPayload,
-            messageType: "StartRecognition"
-          });
-        }
-        ws.send(startPayload);
-      });
-      const onMessage = (data) => {
-        const rawPayload = data.toString();
-        try {
-          const msg = JSON.parse(rawPayload);
-          if (msg.message === "RecognitionStarted") {
-            clearTimeout(timeout);
-            ws.removeListener("message", onMessage);
-            ws.emit("message", data);
-            resolve();
-          } else if (msg.message === "Error") {
-            clearTimeout(timeout);
-            ws.removeListener("message", onMessage);
-            reject(new Error(msg.reason || "Recognition failed to start"));
-          }
-        } catch {
-        }
-      };
-      ws.on("message", onMessage);
-    });
-    ws.on("message", (data) => {
-      const rawPayload = data.toString();
-      try {
-        const message = JSON.parse(rawPayload);
-        if (callbacks?.onRawMessage) {
-          callbacks.onRawMessage({
-            provider: "speechmatics",
-            direction: "incoming",
-            timestamp: Date.now(),
-            payload: rawPayload,
-            messageType: message.message
-          });
-        }
-        this.handleStreamingMessage(message, callbacks, utteranceResults);
-      } catch (error) {
-        if (callbacks?.onRawMessage) {
-          callbacks.onRawMessage({
-            provider: "speechmatics",
-            direction: "incoming",
-            timestamp: Date.now(),
-            payload: rawPayload,
-            messageType: "parse_error"
-          });
-        }
-        callbacks?.onError?.({
-          code: "PARSE_ERROR",
-          message: "Failed to parse WebSocket message",
-          details: error
-        });
-      }
-    });
-    ws.on("error", (error) => {
-      callbacks?.onError?.({
-        code: "WEBSOCKET_ERROR",
-        message: error.message,
-        details: error
-      });
-    });
-    ws.on("close", (code, reason) => {
-      sessionStatus = "closed";
-      callbacks?.onClose?.(code, reason.toString());
-    });
-    await sessionReady;
-    sessionStatus = "open";
-    callbacks?.onOpen?.();
-    return {
-      id: sessionId,
-      provider: this.name,
-      createdAt: /* @__PURE__ */ new Date(),
-      getStatus: () => sessionStatus,
-      sendAudio: async (chunk) => {
-        if (sessionStatus !== "open") {
-          throw new Error(`Cannot send audio: session is ${sessionStatus}`);
-        }
-        if (ws.readyState !== WebSocket6.OPEN) {
-          throw new Error("WebSocket is not open");
-        }
-        if (callbacks?.onRawMessage) {
-          const audioPayload = chunk.data instanceof ArrayBuffer ? chunk.data : chunk.data.buffer.slice(
-            chunk.data.byteOffset,
-            chunk.data.byteOffset + chunk.data.byteLength
-          );
-          callbacks.onRawMessage({
-            provider: this.name,
-            direction: "outgoing",
-            timestamp: Date.now(),
-            payload: audioPayload,
-            messageType: "audio"
-          });
-        }
-        ws.send(chunk.data);
-        seqNo++;
-        if (chunk.isLast) {
-          const endMsg = JSON.stringify({
-            message: "EndOfStream",
-            last_seq_no: seqNo
-          });
-          if (callbacks?.onRawMessage) {
-            callbacks.onRawMessage({
-              provider: this.name,
-              direction: "outgoing",
-              timestamp: Date.now(),
-              payload: endMsg,
-              messageType: "EndOfStream"
-            });
-          }
-          ws.send(endMsg);
-        }
-      },
-      close: async () => {
-        if (sessionStatus === "closed" || sessionStatus === "closing") {
-          return;
-        }
-        sessionStatus = "closing";
-        if (ws.readyState === WebSocket6.OPEN) {
-          seqNo++;
-          ws.send(
-            JSON.stringify({
-              message: "EndOfStream",
-              last_seq_no: seqNo
-            })
-          );
-        }
-        return new Promise((resolve) => {
-          const timeout = setTimeout(() => {
-            ws.terminate();
-            sessionStatus = "closed";
-            resolve();
-          }, 5e3);
-          const onMsg = (data) => {
-            try {
-              const msg = JSON.parse(data.toString());
-              if (msg.message === "EndOfTranscript") {
-                ws.removeListener("message", onMsg);
-                clearTimeout(timeout);
-                ws.close();
-              }
-            } catch {
-            }
-          };
-          ws.on("message", onMsg);
-          ws.once("close", () => {
-            clearTimeout(timeout);
-            sessionStatus = "closed";
-            resolve();
-          });
-        });
-      }
-    };
-  }
-  /**
-   * Handle incoming Speechmatics real-time WebSocket messages
-   */
-  handleStreamingMessage(message, callbacks, utteranceResults) {
-    switch (message.message) {
-      case "RecognitionStarted": {
-        break;
-      }
-      case "AddPartialTranscript": {
-        const results = message.results || [];
-        const text = buildTextFromSpeechmaticsResults(results);
-        if (text) {
-          callbacks?.onTranscript?.({
-            type: "transcript",
-            text,
-            isFinal: false,
-            words: this.extractWordsFromResults(results),
-            data: message
-          });
-        }
-        break;
-      }
-      case "AddTranscript": {
-        const results = message.results || [];
-        const text = buildTextFromSpeechmaticsResults(results);
-        if (utteranceResults) {
-          utteranceResults.push(...results);
-        }
-        if (text) {
-          callbacks?.onTranscript?.({
-            type: "transcript",
-            text,
-            isFinal: true,
-            words: this.extractWordsFromResults(results),
-            data: message
-          });
-        }
-        break;
-      }
-      case "EndOfUtterance": {
-        if (utteranceResults && utteranceResults.length > 0) {
-          const text = buildTextFromSpeechmaticsResults(utteranceResults);
-          const words = this.extractWordsFromResults(utteranceResults);
-          const utterances = buildUtterancesFromWords(words);
-          if (utterances.length > 0) {
-            for (const utt of utterances) {
-              callbacks?.onUtterance?.(utt);
-            }
-          } else if (text) {
-            callbacks?.onUtterance?.({
-              text,
-              start: words.length > 0 ? words[0].start : 0,
-              end: words.length > 0 ? words[words.length - 1].end : 0,
-              words
-            });
-          }
-          utteranceResults.length = 0;
-        }
-        break;
-      }
-      case "AudioAdded": {
-        break;
-      }
-      case "EndOfTranscript": {
-        break;
-      }
-      case "Info":
-      case "Warning": {
-        callbacks?.onMetadata?.(message);
-        break;
-      }
-      case "Error": {
-        const errMsg = message;
-        callbacks?.onError?.({
-          code: errMsg.type || "SPEECHMATICS_ERROR",
-          message: errMsg.reason || "Unknown error",
-          details: message
-        });
-        break;
-      }
-      default: {
-        callbacks?.onMetadata?.(message);
-        break;
-      }
-    }
-  }
-  /**
-   * Extract unified Word[] from Speechmatics recognition results
-   */
-  extractWordsFromResults(results) {
-    return results.filter((r) => r.type === "word" && r.start_time !== void 0 && r.end_time !== void 0).map((result) => ({
-      word: result.alternatives?.[0]?.content || "",
-      start: result.start_time,
-      end: result.end_time,
-      confidence: result.alternatives?.[0]?.confidence,
-      speaker: result.alternatives?.[0]?.speaker
-    }));
-  }
   /**
    * Normalize Speechmatics status to unified status
    * Uses generated JobDetailsStatus enum values
@@ -9441,9 +9099,6 @@ function createSpeechmaticsAdapter(config) {
   return adapter;
 }
 
-// src/adapters/soniox-adapter.ts
-import axios9 from "axios";
-
 // src/generated/soniox/schema/transcriptionStatus.ts
 var TranscriptionStatus = {
   queued: "queued",
@@ -9452,6 +9107,57 @@ var TranscriptionStatus = {
   error: "error"
 };
 
+// src/generated/soniox/api/sonioxPublicAPI.ts
+import axios9 from "axios";
+
+// src/generated/soniox/schema/index.ts
+var schema_exports4 = {};
+__export(schema_exports4, {
+  TemporaryApiKeyUsageType: () => TemporaryApiKeyUsageType,
+  TranscriptionMode: () => TranscriptionMode,
+  TranscriptionStatus: () => TranscriptionStatus,
+  TranslationConfigType: () => TranslationConfigType
+});
+
+// src/generated/soniox/schema/temporaryApiKeyUsageType.ts
+var TemporaryApiKeyUsageType = {
+  transcribe_websocket: "transcribe_websocket"
+};
+
+// src/generated/soniox/schema/transcriptionMode.ts
+var TranscriptionMode = {
+  real_time: "real_time",
+  async: "async"
+};
+
+// src/generated/soniox/schema/translationConfigType.ts
+var TranslationConfigType = {
+  one_way: "one_way",
+  two_way: "two_way"
+};
+
+// src/generated/soniox/api/sonioxPublicAPI.ts
+var uploadFile = (uploadFileBody2, options) => {
+  const formData = new FormData();
+  if (uploadFileBody2.client_reference_id !== void 0 && uploadFileBody2.client_reference_id !== null) {
+    formData.append("client_reference_id", uploadFileBody2.client_reference_id);
+  }
+  formData.append("file", uploadFileBody2.file);
+  return axios9.post("/v1/files", formData, options);
+};
+var createTranscription2 = (createTranscriptionPayload, options) => {
+  return axios9.post("/v1/transcriptions", createTranscriptionPayload, options);
+};
+var getTranscription = (transcriptionId, options) => {
+  return axios9.get(`/v1/transcriptions/${transcriptionId}`, options);
+};
+var getTranscriptionTranscript = (transcriptionId, options) => {
+  return axios9.get(`/v1/transcriptions/${transcriptionId}/transcript`, options);
+};
+var getModels = (options) => {
+  return axios9.get("/v1/models", options);
+};
+
 // src/adapters/soniox-adapter.ts
 var SonioxAdapter = class extends BaseAdapter {
   constructor() {
@@ -9506,11 +9212,17 @@ var SonioxAdapter = class extends BaseAdapter {
     }
   }
   /**
-   * Get the base URL for API requests
+   * Get the base URL for API requests (no /v1 suffix — generated functions include /v1 in paths)
    */
   get baseUrl() {
     if (this.config?.baseUrl) return this.config.baseUrl;
-    return `https://${this.getRegionalHost()}/v1`;
+    return `https://${this.getRegionalHost()}`;
+  }
+  /**
+   * Build axios config with Soniox Bearer auth
+   */
+  getAxiosConfig() {
+    return super.getAxiosConfig("Authorization", (key) => `Bearer ${key}`);
   }
   initialize(config) {
     super.initialize(config);
@@ -9520,15 +9232,6 @@ var SonioxAdapter = class extends BaseAdapter {
     if (config.model) {
       this.defaultModel = config.model;
     }
-    this.client = axios9.create({
-      baseURL: this.baseUrl,
-      timeout: config.timeout || 12e4,
-      headers: {
-        Authorization: `Bearer ${config.apiKey}`,
-        "Content-Type": "application/json",
-        ...config.headers
-      }
-    });
   }
   /**
    * Get current region
@@ -9558,23 +9261,12 @@ var SonioxAdapter = class extends BaseAdapter {
    */
   setRegion(region) {
     this.region = region;
-    if (this.config?.apiKey) {
-      this.client = axios9.create({
-        baseURL: this.baseUrl,
-        timeout: this.config.timeout || 12e4,
-        headers: {
-          Authorization: `Bearer ${this.config.apiKey}`,
-          "Content-Type": "application/json",
-          ...this.config.headers
-        }
-      });
-    }
   }
   /**
    * Submit audio for transcription
    *
-   * Soniox uses async batch processing. The transcribe method submits audio
-   * and waits for completion (or use getTranscript for polling).
+   * Uses the async v1 API: createTranscription returns status `queued`,
+   * then polls until completed (or returns immediately if webhook is set).
    *
    * @param audio - Audio input (URL or file)
    * @param options - Transcription options
@@ -9583,21 +9275,44 @@ var SonioxAdapter = class extends BaseAdapter {
   async transcribe(audio, options) {
     this.validateConfig();
     try {
-      const requestBody = {
-        model: options?.model || this.defaultModel
-      };
-      if (audio.type === "url") {
-        requestBody.audio_url = audio.url;
-      } else if (audio.type === "file") {
-        const formData = new FormData();
+      const sonioxOpts = options?.soniox;
+      if (audio.type === "file") {
         const audioBlob = audio.file instanceof Blob ? audio.file : new Blob([audio.file], { type: audio.mimeType || "audio/wav" });
-        formData.append("file", audioBlob, audio.filename || "audio.wav");
-        const uploadResponse = await this.client.post("/files", formData, {
-          headers: {
-            "Content-Type": "multipart/form-data"
-          }
-        });
-        requestBody.file_id = uploadResponse.data.id;
+        const uploadBody = { file: audioBlob };
+        const fileResp = await uploadFile(uploadBody, this.getAxiosConfig());
+        const payload = {
+          ...sonioxOpts,
+          model: options?.model || this.defaultModel,
+          file_id: fileResp.data.id,
+          language_hints: options?.language ? [options.language] : sonioxOpts?.language_hints,
+          enable_speaker_diarization: options?.diarization || sonioxOpts?.enable_speaker_diarization,
+          enable_language_identification: options?.languageDetection || sonioxOpts?.enable_language_identification,
+          context: options?.customVocabulary?.length ? { terms: options.customVocabulary } : sonioxOpts?.context,
+          webhook_url: options?.webhookUrl || sonioxOpts?.webhook_url
+        };
+        const createResp = await createTranscription2(payload, this.getAxiosConfig());
+        const meta = createResp.data;
+        if (options?.webhookUrl || sonioxOpts?.webhook_url) {
+          return this.normalizeTranscription(meta);
+        }
+        return this.pollForCompletion(meta.id);
+      } else if (audio.type === "url") {
+        const payload = {
+          ...sonioxOpts,
+          model: options?.model || this.defaultModel,
+          audio_url: audio.url,
+          language_hints: options?.language ? [options.language] : sonioxOpts?.language_hints,
+          enable_speaker_diarization: options?.diarization || sonioxOpts?.enable_speaker_diarization,
+          enable_language_identification: options?.languageDetection || sonioxOpts?.enable_language_identification,
+          context: options?.customVocabulary?.length ? { terms: options.customVocabulary } : sonioxOpts?.context,
+          webhook_url: options?.webhookUrl || sonioxOpts?.webhook_url
+        };
+        const createResp = await createTranscription2(payload, this.getAxiosConfig());
+        const meta = createResp.data;
+        if (options?.webhookUrl || sonioxOpts?.webhook_url) {
+          return this.normalizeTranscription(meta);
+        }
+        return this.pollForCompletion(meta.id);
       } else {
         return {
           success: false,
@@ -9608,23 +9323,6 @@ var SonioxAdapter = class extends BaseAdapter {
           }
         };
       }
-      if (options?.language) {
-        requestBody.language_hints = [options.language];
-      }
-      if (options?.diarization) {
-        requestBody.enable_speaker_diarization = true;
-      }
-      if (options?.languageDetection) {
-        requestBody.enable_language_identification = true;
-      }
-      if (options?.customVocabulary && options.customVocabulary.length > 0) {
-        requestBody.context = {
-          terms: options.customVocabulary
-        };
-      }
-      const response = await this.client.post("/transcriptions", requestBody);
-      const transcriptionId = response.data.id;
-      return await this.pollForCompletion(transcriptionId);
     } catch (error) {
       return this.createErrorResponse(error);
     }
@@ -9632,9 +9330,8 @@ var SonioxAdapter = class extends BaseAdapter {
   /**
    * Get transcription result by ID
    *
-   * Checks job status via GET /v1/transcriptions/{id}, then fetches
-   * the full transcript via GET /v1/transcriptions/{id}/transcript
-   * when completed.
+   * Fetches transcription metadata and, if completed, the transcript text/tokens.
+   * Used by pollForCompletion() for async polling.
    *
    * @param transcriptId - Transcript ID
    * @returns Transcription response
@@ -9642,39 +9339,20 @@ var SonioxAdapter = class extends BaseAdapter {
   async getTranscript(transcriptId) {
     this.validateConfig();
     try {
-      const statusResponse = await this.client.get(`/transcriptions/${transcriptId}`);
-      const job = statusResponse.data;
-      if (job.status === "error") {
-        return {
-          success: false,
-          provider: this.name,
-          error: {
-            code: "TRANSCRIPTION_ERROR",
-            message: job.error_message || "Transcription failed"
-          }
-        };
-      }
-      if (job.status !== "completed") {
-        return {
-          success: true,
-          provider: this.name,
-          data: {
-            id: job.id,
-            text: "",
-            status: job.status
-          },
-          raw: job
-        };
+      const metaResp = await getTranscription(transcriptId, this.getAxiosConfig());
+      const meta = metaResp.data;
+      if (meta.status === TranscriptionStatus.completed) {
+        try {
+          const transcriptResp = await getTranscriptionTranscript(
+            transcriptId,
+            this.getAxiosConfig()
+          );
+          return this.normalizeTranscription(meta, transcriptResp.data);
+        } catch (transcriptError) {
+          return this.createErrorResponse(transcriptError);
+        }
       }
-      const transcriptResponse = await this.client.get(
-        `/transcriptions/${transcriptId}/transcript`
-      );
-      return this.normalizeResponse({
-        ...transcriptResponse.data,
-        // Carry over job metadata
-        id: job.id,
-        audio_duration_ms: job.audio_duration_ms
-      });
+      return this.normalizeTranscription(meta);
     } catch (error) {
       return this.createErrorResponse(error);
     }
@@ -9694,51 +9372,50 @@ var SonioxAdapter = class extends BaseAdapter {
     const sessionId = `soniox_${Date.now()}_${Math.random().toString(36).substring(7)}`;
     const createdAt = /* @__PURE__ */ new Date();
     const wsBase = this.config?.wsBaseUrl || (this.config?.baseUrl ? this.deriveWsUrl(this.config.baseUrl) : `wss://${this.getRegionalWsHost()}`);
-    const wsUrl = `${wsBase}/transcribe-websocket`;
-    const modelId = options?.sonioxStreaming?.model || options?.model || "stt-rt-v4";
-    const sonioxOpts = options?.sonioxStreaming;
-    const initMessage = {
-      api_key: this.config.apiKey,
-      model: modelId
-    };
-    if (sonioxOpts?.audioFormat) {
-      initMessage.audio_format = sonioxOpts.audioFormat;
-    } else if (options?.encoding) {
+    const wsUrl = new URL(`${wsBase}/transcribe-websocket`);
+    wsUrl.searchParams.set("api_key", this.config.apiKey);
+    const modelId = options?.sonioxStreaming?.model || options?.model || "stt-rt-preview";
+    wsUrl.searchParams.set("model", modelId);
+    if (options?.encoding) {
       const encodingMap = {
         linear16: "pcm_s16le",
         pcm: "pcm_s16le",
         mulaw: "mulaw",
         alaw: "alaw"
       };
-      initMessage.audio_format = encodingMap[options.encoding] || options.encoding;
+      wsUrl.searchParams.set("audio_format", encodingMap[options.encoding] || options.encoding);
     }
-    if (sonioxOpts?.sampleRate || options?.sampleRate) {
-      initMessage.sample_rate = sonioxOpts?.sampleRate || options?.sampleRate;
+    if (options?.sampleRate) {
+      wsUrl.searchParams.set("sample_rate", options.sampleRate.toString());
     }
-    if (sonioxOpts?.numChannels || options?.channels) {
-      initMessage.num_channels = sonioxOpts?.numChannels || options?.channels;
+    if (options?.channels) {
+      wsUrl.searchParams.set("num_channels", options.channels.toString());
     }
+    const sonioxOpts = options?.sonioxStreaming;
     if (sonioxOpts) {
       if (sonioxOpts.languageHints && sonioxOpts.languageHints.length > 0) {
-        initMessage.language_hints = sonioxOpts.languageHints;
+        wsUrl.searchParams.set("language_hints", JSON.stringify(sonioxOpts.languageHints));
       }
       if (sonioxOpts.enableLanguageIdentification) {
-        initMessage.enable_language_identification = true;
+        wsUrl.searchParams.set("enable_language_identification", "true");
       }
       if (sonioxOpts.enableEndpointDetection) {
-        initMessage.enable_endpoint_detection = true;
+        wsUrl.searchParams.set("enable_endpoint_detection", "true");
       }
       if (sonioxOpts.enableSpeakerDiarization) {
-        initMessage.enable_speaker_diarization = true;
+        wsUrl.searchParams.set("enable_speaker_diarization", "true");
       }
       if (sonioxOpts.context) {
-        initMessage.context = typeof sonioxOpts.context === "string" ? sonioxOpts.context : sonioxOpts.context;
+        wsUrl.searchParams.set(
+          "context",
+          typeof sonioxOpts.context === "string" ? sonioxOpts.context : JSON.stringify(sonioxOpts.context)
+        );
       }
       if (sonioxOpts.translation) {
-        initMessage.translation = sonioxOpts.translation;
+        wsUrl.searchParams.set("translation", JSON.stringify(sonioxOpts.translation));
       }
       if (sonioxOpts.clientReferenceId) {
-        initMessage.client_reference_id = sonioxOpts.clientReferenceId;
+        wsUrl.searchParams.set("client_reference_id", sonioxOpts.clientReferenceId);
       }
     }
     if (!sonioxOpts?.languageHints && options?.language) {
@@ -9747,33 +9424,24 @@ var SonioxAdapter = class extends BaseAdapter {
           `[Soniox] Warning: language="multi" is Deepgram-specific and not supported by Soniox. For automatic language detection, use languageDetection: true instead, or specify a language code like 'en'.`
         );
       }
-      initMessage.language_hints = [options.language];
+      wsUrl.searchParams.set("language_hints", JSON.stringify([options.language]));
     }
     if (!sonioxOpts?.enableSpeakerDiarization && options?.diarization) {
-      initMessage.enable_speaker_diarization = true;
+      wsUrl.searchParams.set("enable_speaker_diarization", "true");
     }
     if (!sonioxOpts?.enableLanguageIdentification && options?.languageDetection) {
-      initMessage.enable_language_identification = true;
+      wsUrl.searchParams.set("enable_language_identification", "true");
+    }
+    if (options?.interimResults !== false) {
     }
     let status = "connecting";
     let openedAt = null;
     let receivedData = false;
     const WebSocketImpl = typeof WebSocket !== "undefined" ? WebSocket : __require("ws");
-    const ws = new WebSocketImpl(wsUrl);
+    const ws = new WebSocketImpl(wsUrl.toString());
     ws.onopen = () => {
-      openedAt = Date.now();
-      const initPayload = JSON.stringify(initMessage);
-      if (callbacks?.onRawMessage) {
-        callbacks.onRawMessage({
-          provider: this.name,
-          direction: "outgoing",
-          timestamp: Date.now(),
-          payload: initPayload,
-          messageType: "init"
-        });
-      }
-      ws.send(initPayload);
       status = "open";
+      openedAt = Date.now();
       callbacks?.onOpen?.();
     };
     ws.onmessage = (event) => {
@@ -9782,7 +9450,8 @@ var SonioxAdapter = class extends BaseAdapter {
       let messageType;
       try {
         const data = JSON.parse(rawPayload);
-        if (data.error) {
+        const errorMessage = data.error_message || data.error;
+        if (errorMessage) {
           messageType = "error";
         } else if (data.finished) {
           messageType = "finished";
@@ -9798,10 +9467,10 @@ var SonioxAdapter = class extends BaseAdapter {
             messageType
           });
         }
-        if (data.error) {
+        if (errorMessage) {
           callbacks?.onError?.({
             code: data.error_code?.toString() || "STREAM_ERROR",
-            message: data.error
+            message: errorMessage
           });
           return;
         }
@@ -9815,7 +9484,7 @@ var SonioxAdapter = class extends BaseAdapter {
             start: token.start_ms ? token.start_ms / 1e3 : 0,
             end: token.end_ms ? token.end_ms / 1e3 : 0,
             confidence: token.confidence,
-            speaker: token.speaker
+            speaker: token.speaker ?? void 0
           }));
           const text = data.text || data.tokens.map((t) => t.text).join("");
           const isFinal = data.tokens.every((t) => t.is_final);
@@ -9824,8 +9493,8 @@ var SonioxAdapter = class extends BaseAdapter {
             text,
             isFinal,
             words,
-            speaker: data.tokens[0]?.speaker,
-            language: data.tokens[0]?.language,
+            speaker: data.tokens[0]?.speaker ?? void 0,
+            language: data.tokens[0]?.language ?? void 0,
             confidence: data.tokens[0]?.confidence
           };
           callbacks?.onTranscript?.(event2);
@@ -9852,10 +9521,10 @@ var SonioxAdapter = class extends BaseAdapter {
     ws.onclose = (event) => {
       status = "closed";
       const timeSinceOpen = openedAt ? Date.now() - openedAt : null;
-      const isEarlyClose = timeSinceOpen !== null && timeSinceOpen < 5e3 && !receivedData;
-      if (isEarlyClose && event.code === 1e3) {
+      const isImmediateClose = timeSinceOpen !== null && timeSinceOpen < 1e3 && !receivedData;
+      if (isImmediateClose && event.code === 1e3) {
         const errorMessage = [
-          "Soniox closed connection shortly after opening.",
+          "Soniox closed connection immediately after opening.",
           `Current config: region=${this.region}, model=${modelId}`,
           "Likely causes:",
           "  - Invalid API key or region mismatch (keys are region-specific, current: " + this.region + ")",
@@ -9941,7 +9610,7 @@ var SonioxAdapter = class extends BaseAdapter {
   async getModels() {
     this.validateConfig();
     try {
-      const response = await this.client.get("/models");
+      const response = await getModels(this.getAxiosConfig());
       return response.data.models || [];
     } catch (error) {
       console.error("Failed to fetch Soniox models:", error);
@@ -9968,55 +9637,82 @@ var SonioxAdapter = class extends BaseAdapter {
       start: token.start_ms ? token.start_ms / 1e3 : 0,
       end: token.end_ms ? token.end_ms / 1e3 : 0,
       confidence: token.confidence,
-      speaker: token.speaker
+      speaker: token.speaker ?? void 0
     }));
     return buildUtterancesFromWords(words);
   }
   /**
-   * Normalize Soniox response to unified format
+   * Normalize v1 API response to unified format
+   *
+   * @param meta - Transcription metadata from getTranscription/createTranscription
+   * @param transcript - Transcript data (text/tokens), only present when status is completed
    */
-  normalizeResponse(response) {
-    const text = response.text || (response.tokens ? response.tokens.filter((t) => t.is_final !== false).map((t) => t.text).join("") : "");
-    const words = response.tokens ? response.tokens.filter(
-      (t) => t.is_final !== false && t.start_ms !== void 0 && t.end_ms !== void 0
-    ).map((token) => ({
+  normalizeTranscription(meta, transcript) {
+    if (meta.status === TranscriptionStatus.error) {
+      return {
+        success: false,
+        provider: this.name,
+        data: {
+          id: meta.id,
+          text: "",
+          status: "error"
+        },
+        error: {
+          code: meta.error_type || "TRANSCRIPTION_ERROR",
+          message: meta.error_message || "Transcription failed"
+        },
+        raw: { meta, transcript }
+      };
+    }
+    if (!transcript) {
+      return {
+        success: true,
+        provider: this.name,
+        data: {
+          id: meta.id,
+          text: "",
+          status: meta.status,
+          duration: meta.audio_duration_ms ? meta.audio_duration_ms / 1e3 : void 0
+        },
+        raw: { meta }
+      };
+    }
+    const tokens = transcript.tokens || [];
+    const text = transcript.text || tokens.map((t) => t.text).join("");
+    const words = tokens.filter((t) => t.start_ms !== void 0 && t.end_ms !== void 0).map((token) => ({
       word: token.text,
       start: token.start_ms / 1e3,
       end: token.end_ms / 1e3,
       confidence: token.confidence,
-      speaker: token.speaker
-    })) : [];
+      speaker: token.speaker ?? void 0
+    }));
     const speakerSet = /* @__PURE__ */ new Set();
-    if (response.tokens) {
-      response.tokens.forEach((t) => {
-        if (t.speaker) speakerSet.add(t.speaker);
-      });
-    }
+    tokens.forEach((t) => {
+      if (t.speaker) speakerSet.add(String(t.speaker));
+    });
     const speakers = speakerSet.size > 0 ? Array.from(speakerSet).map((id) => ({
       id,
       label: `Speaker ${id}`
     })) : void 0;
-    const tokens = response.tokens ? response.tokens.filter((t) => t.is_final !== false) : [];
-    const utterances = tokens.length > 0 ? this.buildUtterancesFromTokens(tokens) : [];
-    const language = response.tokens?.find((t) => t.language)?.language;
+    const utterances = this.buildUtterancesFromTokens(tokens);
+    const language = tokens.find((t) => t.language)?.language ?? void 0;
     return {
       success: true,
       provider: this.name,
       data: {
-        id: response.id || `soniox_${Date.now()}`,
+        id: meta.id,
         text,
         status: TranscriptionStatus.completed,
         language,
-        duration: response.audio_duration_ms ? response.audio_duration_ms / 1e3 : response.total_audio_proc_ms ? response.total_audio_proc_ms / 1e3 : void 0,
+        duration: meta.audio_duration_ms ? meta.audio_duration_ms / 1e3 : void 0,
         speakers,
         words: words.length > 0 ? words : void 0,
         utterances: utterances.length > 0 ? utterances : void 0
       },
       tracking: {
-        requestId: response.id,
-        processingTimeMs: response.total_audio_proc_ms
+        requestId: meta.id
       },
-      raw: response
+      raw: { meta, transcript }
     };
   }
 };
@@ -10440,7 +10136,7 @@ var ElevenLabsAdapter = class extends BaseAdapter {
    * - Multi-channel: `MultichannelSpeechToTextResponseModel` with `transcripts[]`
    */
   normalizeResponse(response) {
-    const chunks = response.transcripts ? response.transcripts : [response];
+    const chunks = "transcripts" in response ? response.transcripts : [response];
     const text = chunks.map((c) => c.text).join(" ");
     const words = [];
     const speakerSet = /* @__PURE__ */ new Set();
@@ -10834,9 +10530,7 @@ var listenTranscribeQueryParams = zod.object({
   ),
   dictation: zod.boolean().optional().describe("Dictation mode for controlling formatting with dictated speech"),
   encoding: zod.enum(["linear16", "flac", "mulaw", "amr-nb", "amr-wb", "opus", "speex", "g729"]).optional().describe("Specify the expected encoding of your submitted audio"),
-  filler_words: zod.boolean().optional().describe(
-    'Filler Words can help transcribe interruptions in your audio, like "uh" and "um"'
-  ),
+  filler_words: zod.boolean().optional().describe('Filler Words can help transcribe interruptions in your audio, like "uh" and "um"'),
   keyterm: zod.array(zod.string()).optional().describe(
     "Key term prompting can boost or suppress specialized terminology and brands. Only compatible with Nova-3"
   ),
@@ -11540,6 +11234,7 @@ __export(assemblyAIAPI_zod_exports, {
   createTranscriptBodySpeechUnderstandingRequestTranslationFormalDefault: () => createTranscriptBodySpeechUnderstandingRequestTranslationFormalDefault,
   createTranscriptBodySpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault: () => createTranscriptBodySpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault,
   createTranscriptBodySummarizationDefault: () => createTranscriptBodySummarizationDefault,
+  createTranscriptBodyTemperatureDefault: () => createTranscriptBodyTemperatureDefault,
   createTranscriptResponse: () => createTranscriptResponse,
   createTranscriptResponseLanguageDetectionOptionsCodeSwitchingConfidenceThresholdDefault: () => createTranscriptResponseLanguageDetectionOptionsCodeSwitchingConfidenceThresholdDefault,
   createTranscriptResponseLanguageDetectionOptionsCodeSwitchingDefault: () => createTranscriptResponseLanguageDetectionOptionsCodeSwitchingDefault,
@@ -11609,6 +11304,7 @@ var createTranscriptBodySpeakerOptionsMinSpeakersExpectedDefault = 1;
 var createTranscriptBodySpeechUnderstandingRequestTranslationFormalDefault = true;
 var createTranscriptBodySpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault = false;
 var createTranscriptBodySummarizationDefault = false;
+var createTranscriptBodyTemperatureDefault = 0;
 var createTranscriptBodyCustomTopicsDefault = false;
 var createTranscriptBody = zod3.object({
   audio_end_at: zod3.number().optional().describe(
@@ -11618,10 +11314,10 @@ var createTranscriptBody = zod3.object({
     "The point in time, in milliseconds, to begin transcribing in your media file. See [Set the start and end of the transcript](https://www.assemblyai.com/docs/pre-recorded-audio/set-the-start-and-end-of-the-transcript) for more details."
   ),
   auto_chapters: zod3.boolean().optional().describe(
-    "Enable [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters), can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Enable [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters), can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   auto_highlights: zod3.boolean().optional().describe(
-    "Enable [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases), either true or false"
+    "Enable [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights), either true or false"
   ),
   content_safety: zod3.boolean().optional().describe(
     "Enable [Content Moderation](https://www.assemblyai.com/docs/content-moderation), can be true or false"
@@ -11637,16 +11333,16 @@ var createTranscriptBody = zod3.object({
       "Object containing words or phrases to replace, and the word or phrase to replace with"
     )
   ).optional().describe(
-    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/custom-spelling) for more details."
+    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/correct-spelling-of-terms) for more details."
   ),
   disfluencies: zod3.boolean().optional().describe(
-    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/filler-words), like "umm", in your media file; can be true or false'
+    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/include-filler-words), like "umm", in your media file; can be true or false'
   ),
   domain: zod3.string().nullish().describe(
     'Enable domain-specific transcription models to improve accuracy for specialized terminology. Set to `"medical-v1"` to enable [Medical Mode](https://www.assemblyai.com/docs/pre-recorded-audio/medical-mode) for improved accuracy of medical terms such as medications, procedures, conditions, and dosages.\n\nSupported languages: English (`en`), Spanish (`es`), German (`de`), French (`fr`). If used with an unsupported language, the parameter is ignored and a warning is returned.\n'
   ),
   entity_detection: zod3.boolean().optional().describe(
-    "Enable [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection), can be true or false"
+    "Enable [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript), can be true or false"
   ),
   filter_profanity: zod3.boolean().optional().describe(
     "Filter profanity from the transcribed text, can be true or false. See [Profanity Filtering](https://www.assemblyai.com/docs/profanity-filtering) for more details."
@@ -11655,7 +11351,7 @@ var createTranscriptBody = zod3.object({
     "Enable [Text Formatting](https://www.assemblyai.com/docs/pre-recorded-audio), can be true or false"
   ),
   iab_categories: zod3.boolean().optional().describe(
-    "Enable [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection), can be true or false"
+    "Enable [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics), can be true or false"
   ),
   keyterms_prompt: zod3.array(zod3.string()).optional().describe(
     "Improve accuracy with up to 200 (for Universal-2) or 1000 (for Universal-3 Pro) domain-specific words or phrases (maximum 6 words per phrase). See [Keyterms Prompting](https://www.assemblyai.com/docs/pre-recorded-audio/keyterms-prompting) for more details.\n"
@@ -11903,7 +11599,7 @@ var createTranscriptBody = zod3.object({
     "Specify options for [Automatic Language Detection](https://www.assemblyai.com/docs/pre-recorded-audio/language-detection)."
   ),
   multichannel: zod3.boolean().optional().describe(
-    "Enable [Multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) transcription, can be true or false."
+    "Enable [Multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) transcription, can be true or false."
   ),
   prompt: zod3.string().optional().describe(
     "Provide natural language prompting of up to 1,500 words of contextual information to the model. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for best practices.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
@@ -11986,23 +11682,23 @@ var createTranscriptBody = zod3.object({
     "The replacement logic for detected PII, can be `entity_type` or `hash`. See [PII redaction](https://www.assemblyai.com/docs/pii-redaction) for more details."
   ),
   sentiment_analysis: zod3.boolean().optional().describe(
-    "Enable [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis), can be true or false"
+    "Enable [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech), can be true or false"
   ),
   speaker_labels: zod3.boolean().optional().describe(
-    "Enable [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization), can be true or false"
+    "Enable [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers), can be true or false"
   ),
   speaker_options: zod3.object({
     min_speakers_expected: zod3.number().default(createTranscriptBodySpeakerOptionsMinSpeakersExpectedDefault).describe(
-      "The minimum number of speakers expected in the audio file. See [Set a range of possible speakers](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-a-range-of-possible-speakers) for more details."
+      "The minimum number of speakers expected in the audio file. See [Set a range of possible speakers](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-a-range-of-possible-speakers) for more details."
     ),
     max_speakers_expected: zod3.number().optional().describe(
-      "<Warning>Setting this parameter too high may hurt model accuracy</Warning>\nThe maximum number of speakers expected in the audio file. The default depends on audio duration: no limit for 0-2 minutes, 10 for 2-10 minutes, and 30 for 10+ minutes. See [Set a range of possible speakers](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-a-range-of-possible-speakers) for more details.\n"
+      "<Warning>Setting this parameter too high may hurt model accuracy</Warning>\nThe maximum number of speakers expected in the audio file. The default depends on audio duration: no limit for 0-2 minutes, 10 for 2-10 minutes, and 30 for 10+ minutes. See [Set a range of possible speakers](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-a-range-of-possible-speakers) for more details.\n"
     )
   }).optional().describe(
-    "Specify options for [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-a-range-of-possible-speakers). Use this to set a range of possible speakers."
+    "Specify options for [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-a-range-of-possible-speakers). Use this to set a range of possible speakers."
   ),
   speakers_expected: zod3.number().nullish().describe(
-    "Tells the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-number-of-speakers-expected) for more details."
+    "Tells the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-number-of-speakers-expected) for more details."
   ),
   speech_models: zod3.array(
     zod3.string().describe(
@@ -12078,7 +11774,7 @@ var createTranscriptBody = zod3.object({
     "Enable speech understanding tasks like [Translation](https://www.assemblyai.com/docs/speech-understanding/translation), [Speaker Identification](https://www.assemblyai.com/docs/speech-understanding/speaker-identification), and [Custom Formatting](https://www.assemblyai.com/docs/speech-understanding/custom-formatting). See the task-specific docs for available options and configuration.\n"
   ),
   summarization: zod3.boolean().optional().describe(
-    "Enable [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization), can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Enable [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts), can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   summary_model: zod3.enum(["informative", "conversational", "catchy"]).optional().describe("The model to summarize the transcript"),
   summary_type: zod3.enum(["bullets", "bullets_verbose", "gist", "headline", "paragraph"]).optional().describe("The type of summary"),
@@ -12087,6 +11783,9 @@ var createTranscriptBody = zod3.object({
   ).or(zod3.null()).optional().describe(
     'Remove [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) from the transcript text. Set to `"all"` to remove all audio tags.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n'
   ),
+  temperature: zod3.number().optional().describe(
+    "Control the amount of randomness injected into the model's response. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for more details.\n\nNote: This parameter can only be used with the Universal-3 Pro model.\n"
+  ),
   webhook_auth_header_name: zod3.string().nullish().describe(
     "The header name to be sent with the transcript completed or failed [webhook](https://www.assemblyai.com/docs/deployment/webhooks-for-pre-recorded-audio) requests"
   ),
@@ -12108,7 +11807,7 @@ var createTranscriptResponseSpeechUnderstandingRequestTranslationFormalDefault =
 var createTranscriptResponseSpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault = false;
 var createTranscriptResponse = zod3.object({
   audio_channels: zod3.number().optional().describe(
-    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) is enabled."
+    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) is enabled."
   ),
   audio_duration: zod3.number().nullish().describe("The duration of this transcript object's media file, in seconds"),
   audio_end_at: zod3.number().nullish().describe(
@@ -12119,10 +11818,10 @@ var createTranscriptResponse = zod3.object({
   ),
   audio_url: zod3.string().describe("The URL of the media that was transcribed"),
   auto_chapters: zod3.boolean().nullish().describe(
-    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   auto_highlights: zod3.boolean().describe(
-    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) is enabled, either true or false"
+    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) is enabled, either true or false"
   ),
   auto_highlights_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -12142,9 +11841,9 @@ var createTranscriptResponse = zod3.object({
       })
     ).describe("A temporally-sequential array of Key Phrases")
   }).describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ),
   chapters: zod3.array(
     zod3.object({
@@ -12157,7 +11856,7 @@ var createTranscriptResponse = zod3.object({
       end: zod3.number().describe("The starting time, in milliseconds, for the chapter")
     }).describe("Chapter of the audio file")
   ).nullish().describe(
-    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for more information."
+    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for more information."
   ),
   confidence: zod3.number().nullish().describe(
     "The confidence score for the transcript, between 0.0 (low confidence) and 1.0 (high confidence)"
@@ -12213,10 +11912,10 @@ var createTranscriptResponse = zod3.object({
       "Object containing words or phrases to replace, and the word or phrase to replace with"
     )
   ).nullish().describe(
-    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/custom-spelling) for more details."
+    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/correct-spelling-of-terms) for more details."
   ),
   disfluencies: zod3.boolean().nullish().describe(
-    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/filler-words), like "umm", in your media file; can be true or false'
+    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/include-filler-words), like "umm", in your media file; can be true or false'
   ),
   domain: zod3.string().nullish().describe(
     'The domain-specific model applied to the transcript. When set to `"medical-v1"`, [Medical Mode](https://www.assemblyai.com/docs/pre-recorded-audio/medical-mode) was used to improve accuracy for medical terminology.\n'
@@ -12278,10 +11977,10 @@ var createTranscriptResponse = zod3.object({
       )
     }).describe("A detected entity")
   ).nullish().describe(
-    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) for more information.\n"
+    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) for more information.\n"
   ),
   entity_detection: zod3.boolean().nullish().describe(
-    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) is enabled, can be true or false"
+    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) is enabled, can be true or false"
   ),
   error: zod3.string().optional().describe("Error message of why the transcript failed"),
   filter_profanity: zod3.boolean().nullish().describe(
@@ -12291,7 +11990,7 @@ var createTranscriptResponse = zod3.object({
     "Whether [Text Formatting](https://www.assemblyai.com/docs/pre-recorded-audio) is enabled, either true or false"
   ),
   iab_categories: zod3.boolean().nullish().describe(
-    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) is enabled, can be true or false"
+    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) is enabled, can be true or false"
   ),
   iab_categories_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -12314,9 +12013,9 @@ var createTranscriptResponse = zod3.object({
     ).describe("An array of results for the Topic Detection model"),
     summary: zod3.record(zod3.string(), zod3.number()).describe("The overall relevance of topic to the entire audio file")
   }).describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ),
   id: zod3.string().uuid().describe("The unique identifier of your transcript"),
   keyterms_prompt: zod3.array(zod3.string()).optional().describe(
@@ -12566,7 +12265,7 @@ var createTranscriptResponse = zod3.object({
     "Specify options for [Automatic Language Detection](https://www.assemblyai.com/docs/pre-recorded-audio/language-detection)."
   ),
   multichannel: zod3.boolean().nullish().describe(
-    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) was enabled in the transcription request, either true or false"
+    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) was enabled in the transcription request, either true or false"
   ),
   prompt: zod3.string().optional().describe(
     "Provide natural language prompting of up to 1,500 words of contextual information to the model. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for best practices.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
@@ -12649,7 +12348,7 @@ var createTranscriptResponse = zod3.object({
     "The replacement logic for detected PII, can be `entity_name` or `hash`. See [PII redaction](https://www.assemblyai.com/docs/pii-redaction) for more details."
   ),
   sentiment_analysis: zod3.boolean().nullish().describe(
-    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) is enabled, can be true or false"
+    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) is enabled, can be true or false"
   ),
   sentiment_analysis_results: zod3.array(
     zod3.object({
@@ -12664,17 +12363,17 @@ var createTranscriptResponse = zod3.object({
         "The channel of this utterance. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     }).describe("The result of the Sentiment Analysis model")
   ).nullish().describe(
-    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) for more information.\n"
+    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) for more information.\n"
   ),
   speaker_labels: zod3.boolean().nullish().describe(
-    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, can be true or false"
+    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, can be true or false"
   ),
   speakers_expected: zod3.number().nullish().describe(
-    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-number-of-speakers-expected) for more details."
+    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-number-of-speakers-expected) for more details."
   ),
   speech_model_used: zod3.string().optional().describe(
     "The speech model to use for the transcription. See [Model Selection](https://www.assemblyai.com/docs/pre-recorded-audio/select-the-speech-model) for available models."
@@ -12777,22 +12476,25 @@ var createTranscriptResponse = zod3.object({
     "The status of your transcript. Possible values are queued, processing, completed, or error."
   ),
   summarization: zod3.boolean().describe(
-    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   summary: zod3.string().nullish().describe(
-    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   summary_model: zod3.string().nullish().describe(
-    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n"
+    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n"
   ),
   summary_type: zod3.string().nullish().describe(
-    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   remove_audio_tags: zod3.enum(["all"]).describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ).or(zod3.null()).optional().describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ),
+  temperature: zod3.number().nullish().describe(
+    "The temperature that was used for the model's response. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for more details.\n\nNote: This parameter can only be used with the Universal-3 Pro model.\n"
+  ),
   text: zod3.string().nullish().describe("The textual transcript of your media file"),
   throttled: zod3.boolean().nullish().describe(
     "True while a request is throttled and false when a request is no longer throttled"
@@ -12813,7 +12515,7 @@ var createTranscriptResponse = zod3.object({
             "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
           ),
           speaker: zod3.string().nullable().describe(
-            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
           )
         })
       ).describe("The words in the utterance."),
@@ -12828,7 +12530,7 @@ var createTranscriptResponse = zod3.object({
       )
     })
   ).nullish().describe(
-    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) for more information.\n"
+    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) for more information.\n"
   ),
   webhook_auth: zod3.boolean().describe(
     "Whether [webhook](https://www.assemblyai.com/docs/deployment/webhooks-for-pre-recorded-audio) authentication details were provided"
@@ -12852,7 +12554,7 @@ var createTranscriptResponse = zod3.object({
         "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     })
   ).nullish().describe(
@@ -12925,7 +12627,7 @@ var getTranscriptResponseSpeechUnderstandingRequestTranslationFormalDefault = tr
 var getTranscriptResponseSpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault = false;
 var getTranscriptResponse = zod3.object({
   audio_channels: zod3.number().optional().describe(
-    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) is enabled."
+    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) is enabled."
   ),
   audio_duration: zod3.number().nullish().describe("The duration of this transcript object's media file, in seconds"),
   audio_end_at: zod3.number().nullish().describe(
@@ -12936,10 +12638,10 @@ var getTranscriptResponse = zod3.object({
   ),
   audio_url: zod3.string().describe("The URL of the media that was transcribed"),
   auto_chapters: zod3.boolean().nullish().describe(
-    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   auto_highlights: zod3.boolean().describe(
-    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) is enabled, either true or false"
+    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) is enabled, either true or false"
   ),
   auto_highlights_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -12959,9 +12661,9 @@ var getTranscriptResponse = zod3.object({
       })
     ).describe("A temporally-sequential array of Key Phrases")
   }).describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ),
   chapters: zod3.array(
     zod3.object({
@@ -12974,7 +12676,7 @@ var getTranscriptResponse = zod3.object({
       end: zod3.number().describe("The starting time, in milliseconds, for the chapter")
     }).describe("Chapter of the audio file")
   ).nullish().describe(
-    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for more information."
+    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for more information."
   ),
   confidence: zod3.number().nullish().describe(
     "The confidence score for the transcript, between 0.0 (low confidence) and 1.0 (high confidence)"
@@ -13030,10 +12732,10 @@ var getTranscriptResponse = zod3.object({
       "Object containing words or phrases to replace, and the word or phrase to replace with"
     )
   ).nullish().describe(
-    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/custom-spelling) for more details."
+    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/correct-spelling-of-terms) for more details."
   ),
   disfluencies: zod3.boolean().nullish().describe(
-    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/filler-words), like "umm", in your media file; can be true or false'
+    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/include-filler-words), like "umm", in your media file; can be true or false'
   ),
   domain: zod3.string().nullish().describe(
     'The domain-specific model applied to the transcript. When set to `"medical-v1"`, [Medical Mode](https://www.assemblyai.com/docs/pre-recorded-audio/medical-mode) was used to improve accuracy for medical terminology.\n'
@@ -13095,10 +12797,10 @@ var getTranscriptResponse = zod3.object({
       )
     }).describe("A detected entity")
   ).nullish().describe(
-    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) for more information.\n"
+    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) for more information.\n"
   ),
   entity_detection: zod3.boolean().nullish().describe(
-    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) is enabled, can be true or false"
+    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) is enabled, can be true or false"
   ),
   error: zod3.string().optional().describe("Error message of why the transcript failed"),
   filter_profanity: zod3.boolean().nullish().describe(
@@ -13108,7 +12810,7 @@ var getTranscriptResponse = zod3.object({
     "Whether [Text Formatting](https://www.assemblyai.com/docs/pre-recorded-audio) is enabled, either true or false"
   ),
   iab_categories: zod3.boolean().nullish().describe(
-    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) is enabled, can be true or false"
+    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) is enabled, can be true or false"
   ),
   iab_categories_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -13131,9 +12833,9 @@ var getTranscriptResponse = zod3.object({
     ).describe("An array of results for the Topic Detection model"),
     summary: zod3.record(zod3.string(), zod3.number()).describe("The overall relevance of topic to the entire audio file")
   }).describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ),
   id: zod3.string().uuid().describe("The unique identifier of your transcript"),
   keyterms_prompt: zod3.array(zod3.string()).optional().describe(
@@ -13383,7 +13085,7 @@ var getTranscriptResponse = zod3.object({
     "Specify options for [Automatic Language Detection](https://www.assemblyai.com/docs/pre-recorded-audio/language-detection)."
   ),
   multichannel: zod3.boolean().nullish().describe(
-    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) was enabled in the transcription request, either true or false"
+    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) was enabled in the transcription request, either true or false"
   ),
   prompt: zod3.string().optional().describe(
     "Provide natural language prompting of up to 1,500 words of contextual information to the model. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for best practices.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
@@ -13466,7 +13168,7 @@ var getTranscriptResponse = zod3.object({
     "The replacement logic for detected PII, can be `entity_name` or `hash`. See [PII redaction](https://www.assemblyai.com/docs/pii-redaction) for more details."
   ),
   sentiment_analysis: zod3.boolean().nullish().describe(
-    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) is enabled, can be true or false"
+    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) is enabled, can be true or false"
   ),
   sentiment_analysis_results: zod3.array(
     zod3.object({
@@ -13481,17 +13183,17 @@ var getTranscriptResponse = zod3.object({
         "The channel of this utterance. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     }).describe("The result of the Sentiment Analysis model")
   ).nullish().describe(
-    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) for more information.\n"
+    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) for more information.\n"
   ),
   speaker_labels: zod3.boolean().nullish().describe(
-    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, can be true or false"
+    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, can be true or false"
   ),
   speakers_expected: zod3.number().nullish().describe(
-    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-number-of-speakers-expected) for more details."
+    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-number-of-speakers-expected) for more details."
   ),
   speech_model_used: zod3.string().optional().describe(
     "The speech model to use for the transcription. See [Model Selection](https://www.assemblyai.com/docs/pre-recorded-audio/select-the-speech-model) for available models."
@@ -13594,22 +13296,25 @@ var getTranscriptResponse = zod3.object({
     "The status of your transcript. Possible values are queued, processing, completed, or error."
   ),
   summarization: zod3.boolean().describe(
-    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   summary: zod3.string().nullish().describe(
-    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   summary_model: zod3.string().nullish().describe(
-    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n"
+    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n"
   ),
   summary_type: zod3.string().nullish().describe(
-    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   remove_audio_tags: zod3.enum(["all"]).describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ).or(zod3.null()).optional().describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ),
+  temperature: zod3.number().nullish().describe(
+    "The temperature that was used for the model's response. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for more details.\n\nNote: This parameter can only be used with the Universal-3 Pro model.\n"
+  ),
   text: zod3.string().nullish().describe("The textual transcript of your media file"),
   throttled: zod3.boolean().nullish().describe(
     "True while a request is throttled and false when a request is no longer throttled"
@@ -13630,7 +13335,7 @@ var getTranscriptResponse = zod3.object({
             "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
           ),
           speaker: zod3.string().nullable().describe(
-            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
           )
         })
       ).describe("The words in the utterance."),
@@ -13645,7 +13350,7 @@ var getTranscriptResponse = zod3.object({
       )
     })
   ).nullish().describe(
-    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) for more information.\n"
+    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) for more information.\n"
   ),
   webhook_auth: zod3.boolean().describe(
     "Whether [webhook](https://www.assemblyai.com/docs/deployment/webhooks-for-pre-recorded-audio) authentication details were provided"
@@ -13669,7 +13374,7 @@ var getTranscriptResponse = zod3.object({
         "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     })
   ).nullish().describe(
@@ -13702,7 +13407,7 @@ var deleteTranscriptResponseSpeechUnderstandingRequestTranslationFormalDefault =
 var deleteTranscriptResponseSpeechUnderstandingRequestTranslationMatchOriginalUtteranceDefault = false;
 var deleteTranscriptResponse = zod3.object({
   audio_channels: zod3.number().optional().describe(
-    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) is enabled."
+    "The number of audio channels in the audio file. This is only present when [multichannel](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) is enabled."
   ),
   audio_duration: zod3.number().nullish().describe("The duration of this transcript object's media file, in seconds"),
   audio_end_at: zod3.number().nullish().describe(
@@ -13713,10 +13418,10 @@ var deleteTranscriptResponse = zod3.object({
   ),
   audio_url: zod3.string().describe("The URL of the media that was transcribed"),
   auto_chapters: zod3.boolean().nullish().describe(
-    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) is enabled, can be true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible chapter summaries. See the [updated Auto Chapters page](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   auto_highlights: zod3.boolean().describe(
-    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) is enabled, either true or false"
+    "Whether [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) is enabled, either true or false"
   ),
   auto_highlights_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -13736,9 +13441,9 @@ var deleteTranscriptResponse = zod3.object({
       })
     ).describe("A temporally-sequential array of Key Phrases")
   }).describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/key-phrases) for more information.\n"
+    "An array of results for the Key Phrases model, if it is enabled.\nSee [Key Phrases](https://www.assemblyai.com/docs/speech-understanding/identify-highlights) for more information.\n"
   ),
   chapters: zod3.array(
     zod3.object({
@@ -13751,7 +13456,7 @@ var deleteTranscriptResponse = zod3.object({
       end: zod3.number().describe("The starting time, in milliseconds, for the chapter")
     }).describe("Chapter of the audio file")
   ).nullish().describe(
-    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/auto-chapters) for more information."
+    "An array of temporally sequential chapters for the audio file. See [Auto Chapters](https://www.assemblyai.com/docs/speech-understanding/create-summarized-chapters) for more information."
   ),
   confidence: zod3.number().nullish().describe(
     "The confidence score for the transcript, between 0.0 (low confidence) and 1.0 (high confidence)"
@@ -13807,10 +13512,10 @@ var deleteTranscriptResponse = zod3.object({
       "Object containing words or phrases to replace, and the word or phrase to replace with"
     )
   ).nullish().describe(
-    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/custom-spelling) for more details."
+    "Customize how words are spelled and formatted using to and from values. See [Custom Spelling](https://www.assemblyai.com/docs/pre-recorded-audio/correct-spelling-of-terms) for more details."
   ),
   disfluencies: zod3.boolean().nullish().describe(
-    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/filler-words), like "umm", in your media file; can be true or false'
+    'Transcribe [Filler Words](https://www.assemblyai.com/docs/pre-recorded-audio/include-filler-words), like "umm", in your media file; can be true or false'
   ),
   domain: zod3.string().nullish().describe(
     'The domain-specific model applied to the transcript. When set to `"medical-v1"`, [Medical Mode](https://www.assemblyai.com/docs/pre-recorded-audio/medical-mode) was used to improve accuracy for medical terminology.\n'
@@ -13872,10 +13577,10 @@ var deleteTranscriptResponse = zod3.object({
       )
     }).describe("A detected entity")
   ).nullish().describe(
-    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) for more information.\n"
+    "An array of results for the Entity Detection model, if it is enabled.\nSee [Entity detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) for more information.\n"
   ),
   entity_detection: zod3.boolean().nullish().describe(
-    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/entity-detection) is enabled, can be true or false"
+    "Whether [Entity Detection](https://www.assemblyai.com/docs/speech-understanding/detect-entities-in-transcript) is enabled, can be true or false"
   ),
   error: zod3.string().optional().describe("Error message of why the transcript failed"),
   filter_profanity: zod3.boolean().nullish().describe(
@@ -13885,7 +13590,7 @@ var deleteTranscriptResponse = zod3.object({
     "Whether [Text Formatting](https://www.assemblyai.com/docs/pre-recorded-audio) is enabled, either true or false"
   ),
   iab_categories: zod3.boolean().nullish().describe(
-    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) is enabled, can be true or false"
+    "Whether [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) is enabled, can be true or false"
   ),
   iab_categories_result: zod3.object({
     status: zod3.enum(["success", "unavailable"]).describe("Either success, or unavailable in the rare case that the model failed"),
@@ -13908,9 +13613,9 @@ var deleteTranscriptResponse = zod3.object({
     ).describe("An array of results for the Topic Detection model"),
     summary: zod3.record(zod3.string(), zod3.number()).describe("The overall relevance of topic to the entire audio file")
   }).describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ).or(zod3.null()).optional().describe(
-    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/topic-detection) for more information.\n"
+    "The result of the Topic Detection model, if it is enabled.\nSee [Topic Detection](https://www.assemblyai.com/docs/speech-understanding/detect-discussion-topics) for more information.\n"
   ),
   id: zod3.string().uuid().describe("The unique identifier of your transcript"),
   keyterms_prompt: zod3.array(zod3.string()).optional().describe(
@@ -14160,7 +13865,7 @@ var deleteTranscriptResponse = zod3.object({
     "Specify options for [Automatic Language Detection](https://www.assemblyai.com/docs/pre-recorded-audio/language-detection)."
   ),
   multichannel: zod3.boolean().nullish().describe(
-    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) was enabled in the transcription request, either true or false"
+    "Whether [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) was enabled in the transcription request, either true or false"
   ),
   prompt: zod3.string().optional().describe(
     "Provide natural language prompting of up to 1,500 words of contextual information to the model. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for best practices.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
@@ -14243,7 +13948,7 @@ var deleteTranscriptResponse = zod3.object({
     "The replacement logic for detected PII, can be `entity_name` or `hash`. See [PII redaction](https://www.assemblyai.com/docs/pii-redaction) for more details."
   ),
   sentiment_analysis: zod3.boolean().nullish().describe(
-    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) is enabled, can be true or false"
+    "Whether [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) is enabled, can be true or false"
   ),
   sentiment_analysis_results: zod3.array(
     zod3.object({
@@ -14258,17 +13963,17 @@ var deleteTranscriptResponse = zod3.object({
         "The channel of this utterance. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     }).describe("The result of the Sentiment Analysis model")
   ).nullish().describe(
-    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/sentiment-analysis) for more information.\n"
+    "An array of results for the Sentiment Analysis model, if it is enabled.\nSee [Sentiment Analysis](https://www.assemblyai.com/docs/speech-understanding/analyze-sentiment-of-speech) for more information.\n"
   ),
   speaker_labels: zod3.boolean().nullish().describe(
-    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, can be true or false"
+    "Whether [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, can be true or false"
   ),
   speakers_expected: zod3.number().nullish().describe(
-    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization#set-number-of-speakers-expected) for more details."
+    "Tell the speaker label model how many speakers it should attempt to identify. See [Set number of speakers expected](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers#set-number-of-speakers-expected) for more details."
   ),
   speech_model_used: zod3.string().optional().describe(
     "The speech model to use for the transcription. See [Model Selection](https://www.assemblyai.com/docs/pre-recorded-audio/select-the-speech-model) for available models."
@@ -14371,22 +14076,25 @@ var deleteTranscriptResponse = zod3.object({
     "The status of your transcript. Possible values are queued, processing, completed, or error."
   ),
   summarization: zod3.boolean().describe(
-    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
+    "Whether [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled, either true or false. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n\nNote: This parameter is only supported for the Universal-2 model.\n"
   ),
   summary: zod3.string().nullish().describe(
-    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The generated summary of the media file, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   summary_model: zod3.string().nullish().describe(
-    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details.\n"
+    "The Summarization model used to generate the summary,\nif [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-models) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details.\n"
   ),
   summary_type: zod3.string().nullish().describe(
-    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarization#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarization) for details."
+    "The type of summary generated, if [Summarization](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts#summary-types) is enabled. Deprecated - use [LLM Gateway](https://www.assemblyai.com/docs/llm-gateway/overview) instead for more flexible summaries. See the [updated Summarization page](https://www.assemblyai.com/docs/speech-understanding/summarize-transcripts) for details."
   ),
   remove_audio_tags: zod3.enum(["all"]).describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ).or(zod3.null()).optional().describe(
     "Whether [audio event tags](https://www.assemblyai.com/docs/pre-recorded-audio/universal-3-pro#audio-event-tags) were removed from the transcript text.\n\nNote: This parameter is only supported for the Universal-3 Pro model.\n"
   ),
+  temperature: zod3.number().nullish().describe(
+    "The temperature that was used for the model's response. See the [Prompting Guide](https://www.assemblyai.com/docs/pre-recorded-audio/prompting) for more details.\n\nNote: This parameter can only be used with the Universal-3 Pro model.\n"
+  ),
   text: zod3.string().nullish().describe("The textual transcript of your media file"),
   throttled: zod3.boolean().nullish().describe(
     "True while a request is throttled and false when a request is no longer throttled"
@@ -14407,7 +14115,7 @@ var deleteTranscriptResponse = zod3.object({
             "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
           ),
           speaker: zod3.string().nullable().describe(
-            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
           )
         })
       ).describe("The words in the utterance."),
@@ -14422,7 +14130,7 @@ var deleteTranscriptResponse = zod3.object({
       )
     })
   ).nullish().describe(
-    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/multichannel) for more information.\n"
+    "When multichannel or speaker_labels is enabled, a list of turn-by-turn utterance objects.\nSee [Speaker diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) and [Multichannel transcription](https://www.assemblyai.com/docs/pre-recorded-audio/transcribe-multiple-audio-channels) for more information.\n"
   ),
   webhook_auth: zod3.boolean().describe(
     "Whether [webhook](https://www.assemblyai.com/docs/deployment/webhooks-for-pre-recorded-audio) authentication details were provided"
@@ -14446,7 +14154,7 @@ var deleteTranscriptResponse = zod3.object({
         "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     })
   ).nullish().describe(
@@ -14491,7 +14199,7 @@ var getTranscriptSentencesResponse = zod3.object({
             "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
           ),
           speaker: zod3.string().nullable().describe(
-            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
           )
         })
       ).describe("An array of words in the sentence"),
@@ -14499,7 +14207,7 @@ var getTranscriptSentencesResponse = zod3.object({
         "The channel of the sentence. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
       ),
       speaker: zod3.string().nullable().describe(
-        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+        "The speaker of the sentence if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
       )
     })
   ).describe("An array of sentences in the transcript")
@@ -14527,7 +14235,7 @@ var getTranscriptParagraphsResponse = zod3.object({
             "The channel of the word. The left and right channels are channels 1 and 2. Additional channels increment the channel number sequentially."
           ),
           speaker: zod3.string().nullable().describe(
-            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/speaker-diarization) is enabled, else null"
+            "The speaker of the word if [Speaker Diarization](https://www.assemblyai.com/docs/pre-recorded-audio/label-speakers) is enabled, else null"
           )
         })
       ).describe("An array of words in the paragraph")
@@ -17139,23 +16847,6 @@ var preRecordedControllerGetPreRecordedJobsV2Response = zod5.object({
         }).optional().describe(
           "If `name_consistency` has been enabled, Gladia will improve consistency of the names accross the transcription"
         ),
-        speaker_reidentification: zod5.object({
-          success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
-          is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
-          exec_time: zod5.number().describe("Time audio intelligence model took to complete the task"),
-          error: zod5.object({
-            status_code: zod5.number().describe("Status code of the addon error"),
-            exception: zod5.string().describe("Reason of the addon error"),
-            message: zod5.string().describe("Detailed message of the addon error")
-          }).nullable().describe(
-            "`null` if `success` is `true`. Contains the error details of the failed model"
-          ),
-          results: zod5.string().describe(
-            "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-          )
-        }).optional().describe(
-          "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-        ),
         structured_data_extraction: zod5.object({
           success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
           is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
@@ -18636,23 +18327,6 @@ var preRecordedControllerGetPreRecordedJobV2Response = zod5.object({
     }).optional().describe(
       "If `name_consistency` has been enabled, Gladia will improve consistency of the names accross the transcription"
     ),
-    speaker_reidentification: zod5.object({
-      success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
-      is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
-      exec_time: zod5.number().describe("Time audio intelligence model took to complete the task"),
-      error: zod5.object({
-        status_code: zod5.number().describe("Status code of the addon error"),
-        exception: zod5.string().describe("Reason of the addon error"),
-        message: zod5.string().describe("Detailed message of the addon error")
-      }).nullable().describe(
-        "`null` if `success` is `true`. Contains the error details of the failed model"
-      ),
-      results: zod5.string().describe(
-        "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-      )
-    }).optional().describe(
-      "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-    ),
     structured_data_extraction: zod5.object({
       success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
       is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
@@ -20790,23 +20464,6 @@ var transcriptionControllerListV2Response = zod5.object({
         }).optional().describe(
           "If `name_consistency` has been enabled, Gladia will improve consistency of the names accross the transcription"
         ),
-        speaker_reidentification: zod5.object({
-          success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
-          is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
-          exec_time: zod5.number().describe("Time audio intelligence model took to complete the task"),
-          error: zod5.object({
-            status_code: zod5.number().describe("Status code of the addon error"),
-            exception: zod5.string().describe("Reason of the addon error"),
-            message: zod5.string().describe("Detailed message of the addon error")
-          }).nullable().describe(
-            "`null` if `success` is `true`. Contains the error details of the failed model"
-          ),
-          results: zod5.string().describe(
-            "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-          )
-        }).optional().describe(
-          "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-        ),
         structured_data_extraction: zod5.object({
           success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
           is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
@@ -21106,11 +20763,7 @@ var transcriptionControllerListV2Response = zod5.object({
           channels: zod5.number().min(1).max(transcriptionControllerListV2ResponseItemsItemRequestParamsChannelsMax).default(
             transcriptionControllerListV2ResponseItemsItemRequestParamsChannelsDefault
           ).describe("The number of channels of the audio stream"),
-          model: zod5.enum(["solaria-1"]).describe(
-            'The model used to process the audio. "solaria-1" is used by default.'
-          ).default(transcriptionControllerListV2ResponseItemsItemRequestParamsModelDefault).describe(
-            'The model used to process the audio. "solaria-1" is used by default.'
-          ),
+          model: zod5.enum(["solaria-1"]).describe('The model used to process the audio. "solaria-1" is used by default.').default(transcriptionControllerListV2ResponseItemsItemRequestParamsModelDefault).describe('The model used to process the audio. "solaria-1" is used by default.'),
           endpointing: zod5.number().min(transcriptionControllerListV2ResponseItemsItemRequestParamsEndpointingMin).max(transcriptionControllerListV2ResponseItemsItemRequestParamsEndpointingMax).default(
             transcriptionControllerListV2ResponseItemsItemRequestParamsEndpointingDefault
           ).describe(
@@ -23534,23 +23187,6 @@ var transcriptionControllerGetTranscriptV2Response = zod5.discriminatedUnion("ki
       }).optional().describe(
         "If `name_consistency` has been enabled, Gladia will improve consistency of the names accross the transcription"
       ),
-      speaker_reidentification: zod5.object({
-        success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
-        is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
-        exec_time: zod5.number().describe("Time audio intelligence model took to complete the task"),
-        error: zod5.object({
-          status_code: zod5.number().describe("Status code of the addon error"),
-          exception: zod5.string().describe("Reason of the addon error"),
-          message: zod5.string().describe("Detailed message of the addon error")
-        }).nullable().describe(
-          "`null` if `success` is `true`. Contains the error details of the failed model"
-        ),
-        results: zod5.string().describe(
-          "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-        )
-      }).optional().describe(
-        "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-      ),
       structured_data_extraction: zod5.object({
         success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
         is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
@@ -26716,23 +26352,6 @@ var historyControllerGetListV1Response = zod5.object({
         }).optional().describe(
           "If `name_consistency` has been enabled, Gladia will improve consistency of the names accross the transcription"
         ),
-        speaker_reidentification: zod5.object({
-          success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
-          is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
-          exec_time: zod5.number().describe("Time audio intelligence model took to complete the task"),
-          error: zod5.object({
-            status_code: zod5.number().describe("Status code of the addon error"),
-            exception: zod5.string().describe("Reason of the addon error"),
-            message: zod5.string().describe("Detailed message of the addon error")
-          }).nullable().describe(
-            "`null` if `success` is `true`. Contains the error details of the failed model"
-          ),
-          results: zod5.string().describe(
-            "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-          )
-        }).optional().describe(
-          "If `speaker_reidentification` has been enabled, results of the AI speaker reidentification."
-        ),
         structured_data_extraction: zod5.object({
           success: zod5.boolean().describe("The audio intelligence model succeeded to get a valid output"),
           is_empty: zod5.boolean().describe("The audio intelligence model returned an empty value"),
@@ -27026,11 +26645,7 @@ var historyControllerGetListV1Response = zod5.object({
             historyControllerGetListV1ResponseItemsItemRequestParamsSampleRateDefault
           ).describe("The sample rate of the audio stream"),
           channels: zod5.number().min(1).max(historyControllerGetListV1ResponseItemsItemRequestParamsChannelsMax).default(historyControllerGetListV1ResponseItemsItemRequestParamsChannelsDefault).describe("The number of channels of the audio stream"),
-          model: zod5.enum(["solaria-1"]).describe(
-            'The model used to process the audio. "solaria-1" is used by default.'
-          ).default(historyControllerGetListV1ResponseItemsItemRequestParamsModelDefault).describe(
-            'The model used to process the audio. "solaria-1" is used by default.'
-          ),
+          model: zod5.enum(["solaria-1"]).describe('The model used to process the audio. "solaria-1" is used by default.').default(historyControllerGetListV1ResponseItemsItemRequestParamsModelDefault).describe('The model used to process the audio. "solaria-1" is used by default.'),
           endpointing: zod5.number().min(historyControllerGetListV1ResponseItemsItemRequestParamsEndpointingMin).max(historyControllerGetListV1ResponseItemsItemRequestParamsEndpointingMax).default(
             historyControllerGetListV1ResponseItemsItemRequestParamsEndpointingDefault
           ).describe(
@@ -36191,6 +35806,7 @@ __export(sonioxPublicAPI_zod_exports, {
   createTemporaryApiKeyBody: () => createTemporaryApiKeyBody,
   createTemporaryApiKeyBodyClientReferenceIdMaxOne: () => createTemporaryApiKeyBodyClientReferenceIdMaxOne,
   createTemporaryApiKeyBodyExpiresInSecondsMax: () => createTemporaryApiKeyBodyExpiresInSecondsMax,
+  createTemporaryApiKeyBodyMaxSessionDurationSecondsMaxOne: () => createTemporaryApiKeyBodyMaxSessionDurationSecondsMaxOne,
   createTranscriptionBody: () => createTranscriptionBody2,
   createTranscriptionBodyAudioUrlMaxOne: () => createTranscriptionBodyAudioUrlMaxOne,
   createTranscriptionBodyAudioUrlRegExpOne: () => createTranscriptionBodyAudioUrlRegExpOne,
@@ -36321,11 +35937,11 @@ var getTranscriptionsResponse = zod10.object({
 });
 var createTranscriptionBodyModelMaxThree = 32;
 var createTranscriptionBodyAudioUrlMaxOne = 4096;
-var createTranscriptionBodyAudioUrlRegExpOne = new RegExp("^https?://[^\\s]+$");
+var createTranscriptionBodyAudioUrlRegExpOne = /^https?:\/\/[^\s]+$/;
 var createTranscriptionBodyLanguageHintsItemMax = 10;
 var createTranscriptionBodyLanguageHintsMaxOne = 100;
 var createTranscriptionBodyWebhookUrlMaxOne = 256;
-var createTranscriptionBodyWebhookUrlRegExpOne = new RegExp("^https?://[^\\s]+$");
+var createTranscriptionBodyWebhookUrlRegExpOne = /^https?:\/\/[^\s]+$/;
 var createTranscriptionBodyWebhookAuthHeaderNameMaxOne = 256;
 var createTranscriptionBodyWebhookAuthHeaderValueMaxOne = 256;
 var createTranscriptionBodyClientReferenceIdMaxOne = 256;
@@ -36473,22 +36089,25 @@ var getModelsResponse = zod10.object({
 });
 var createTemporaryApiKeyBodyExpiresInSecondsMax = 3600;
 var createTemporaryApiKeyBodyClientReferenceIdMaxOne = 256;
+var createTemporaryApiKeyBodyMaxSessionDurationSecondsMaxOne = 18e3;
 var createTemporaryApiKeyBody = zod10.object({
   usage_type: zod10.enum(["transcribe_websocket"]),
   expires_in_seconds: zod10.number().min(1).max(createTemporaryApiKeyBodyExpiresInSecondsMax).describe("Duration in seconds until the temporary API key expires."),
-  client_reference_id: zod10.string().max(createTemporaryApiKeyBodyClientReferenceIdMaxOne).or(zod10.null()).optional().describe("Optional tracking identifier string. Does not need to be unique.")
+  client_reference_id: zod10.string().max(createTemporaryApiKeyBodyClientReferenceIdMaxOne).or(zod10.null()).optional().describe("Optional tracking identifier string. Does not need to be unique."),
+  single_use: zod10.boolean().or(zod10.null()).optional().describe("If true, the temporary API key can be used only once."),
+  max_session_duration_seconds: zod10.number().min(1).max(createTemporaryApiKeyBodyMaxSessionDurationSecondsMaxOne).or(zod10.null()).optional().describe(
+    "Maximum WebSocket connection duration in seconds. If exceeded, the connection will be dropped. If not set, no limit is applied."
+  )
 });
 
 // src/generated/soniox/streaming-types.zod.ts
 var streaming_types_zod_exports = {};
 __export(streaming_types_zod_exports, {
   sonioxAudioFormatSchema: () => sonioxAudioFormatSchema,
-  sonioxAutoDetectedAudioFormatSchema: () => sonioxAutoDetectedAudioFormatSchema,
   sonioxContextGeneralItemSchema: () => sonioxContextGeneralItemSchema,
   sonioxContextSchema: () => sonioxContextSchema,
   sonioxErrorStatusSchema: () => sonioxErrorStatusSchema,
   sonioxOneWayTranslationSchema: () => sonioxOneWayTranslationSchema,
-  sonioxPcmAudioEncodingSchema: () => sonioxPcmAudioEncodingSchema,
   sonioxRealtimeModelSchema: () => sonioxRealtimeModelSchema,
   sonioxRecorderStateSchema: () => sonioxRecorderStateSchema,
   sonioxStreamingResponseSchema: () => sonioxStreamingResponseSchema,
@@ -36502,7 +36121,7 @@ __export(streaming_types_zod_exports, {
   streamingUpdateConfigParams: () => streamingUpdateConfigParams3
 });
 import { z as zod11 } from "zod";
-var sonioxAutoDetectedAudioFormatSchema = zod11.enum([
+var sonioxAudioFormatSchema = zod11.enum([
   "auto",
   "aac",
   "aiff",
@@ -36512,10 +36131,7 @@ var sonioxAutoDetectedAudioFormatSchema = zod11.enum([
   "mp3",
   "ogg",
   "wav",
-  "webm"
-]);
-var sonioxPcmAudioEncodingSchema = zod11.enum([
-  // Signed PCM
+  "webm",
   "pcm_s8",
   "pcm_s16le",
   "pcm_s16be",
@@ -36523,7 +36139,6 @@ var sonioxPcmAudioEncodingSchema = zod11.enum([
   "pcm_s24be",
   "pcm_s32le",
   "pcm_s32be",
-  // Unsigned PCM
   "pcm_u8",
   "pcm_u16le",
   "pcm_u16be",
@@ -36531,86 +36146,81 @@ var sonioxPcmAudioEncodingSchema = zod11.enum([
   "pcm_u24be",
   "pcm_u32le",
   "pcm_u32be",
-  // Float PCM
   "pcm_f32le",
   "pcm_f32be",
   "pcm_f64le",
   "pcm_f64be",
-  // Companded
   "mulaw",
   "alaw"
 ]);
-var sonioxAudioFormatSchema = zod11.union([
-  sonioxAutoDetectedAudioFormatSchema,
-  sonioxPcmAudioEncodingSchema
-]);
 var sonioxOneWayTranslationSchema = zod11.object({
   type: zod11.literal("one_way"),
-  target_language: zod11.string().describe("Target language code for translation")
+  target_language: zod11.string()
 });
 var sonioxTwoWayTranslationSchema = zod11.object({
   type: zod11.literal("two_way"),
-  language_a: zod11.string().describe("First language for bidirectional translation"),
-  language_b: zod11.string().describe("Second language for bidirectional translation")
+  language_a: zod11.string(),
+  language_b: zod11.string()
 });
 var sonioxTranslationConfigSchema = zod11.union([
   sonioxOneWayTranslationSchema,
   sonioxTwoWayTranslationSchema
 ]);
 var sonioxContextGeneralItemSchema = zod11.object({
-  key: zod11.string().describe("Context item key (e.g. 'Domain')"),
-  value: zod11.string().describe("Context item value (e.g. 'medicine')")
+  key: zod11.string(),
+  value: zod11.string()
 });
 var sonioxTranslationTermSchema = zod11.object({
-  source: zod11.string().describe("Source term"),
-  target: zod11.string().describe("Target term to translate to")
+  source: zod11.string(),
+  target: zod11.string()
 });
 var sonioxStructuredContextSchema = zod11.object({
-  general: zod11.array(sonioxContextGeneralItemSchema).optional().describe("General context items (key-value pairs)"),
-  text: zod11.string().optional().describe("Text context"),
-  terms: zod11.array(zod11.string()).optional().describe("Terms that might occur in speech"),
-  translation_terms: zod11.array(sonioxTranslationTermSchema).optional().describe("Hints how to translate specific terms (ignored if translation is not enabled)")
+  general: zod11.array(sonioxContextGeneralItemSchema).optional(),
+  text: zod11.string().optional(),
+  terms: zod11.array(zod11.string()).optional(),
+  translation_terms: zod11.array(sonioxTranslationTermSchema).optional()
 });
 var sonioxContextSchema = zod11.union([sonioxStructuredContextSchema, zod11.string()]);
 var sonioxRealtimeModelSchema = zod11.enum([
+  "stt-rt-v4",
   "stt-rt-v3",
   "stt-rt-preview",
   "stt-rt-v3-preview",
   "stt-rt-preview-v2"
 ]);
 var streamingTranscriberParams3 = zod11.object({
-  model: sonioxRealtimeModelSchema.describe("Real-time model to use"),
-  audioFormat: sonioxAudioFormatSchema.optional().describe("Audio format specification. Use 'auto' for automatic detection"),
-  sampleRate: zod11.number().optional().describe("Sample rate in Hz (required for raw PCM formats)"),
-  numChannels: zod11.number().min(1).max(2).optional().describe("Number of audio channels (1 for mono, 2 for stereo) - required for raw PCM formats"),
-  languageHints: zod11.array(zod11.string()).optional().describe("Expected languages in the audio (ISO language codes)"),
-  context: sonioxContextSchema.optional().describe("Additional context to improve transcription accuracy"),
-  enableSpeakerDiarization: zod11.boolean().optional().describe("Enable speaker diarization - each token will include a speaker field"),
-  enableLanguageIdentification: zod11.boolean().optional().describe("Enable language identification - each token will include a language field"),
-  enableEndpointDetection: zod11.boolean().optional().describe("Enable endpoint detection to detect when a speaker has finished talking"),
-  translation: sonioxTranslationConfigSchema.optional().describe("Translation configuration"),
-  clientReferenceId: zod11.string().optional().describe("Optional tracking identifier (client-defined)")
-});
-var sonioxTranslationStatusSchema = zod11.enum(["none", "original", "translation"]);
+  model: sonioxRealtimeModelSchema,
+  audioFormat: sonioxAudioFormatSchema.optional(),
+  sampleRate: zod11.number().optional(),
+  numChannels: zod11.number().optional(),
+  languageHints: zod11.array(zod11.string()).optional(),
+  context: sonioxContextSchema.optional(),
+  enableSpeakerDiarization: zod11.boolean().optional(),
+  enableLanguageIdentification: zod11.boolean().optional(),
+  enableEndpointDetection: zod11.boolean().optional(),
+  translation: sonioxTranslationConfigSchema.optional(),
+  clientReferenceId: zod11.string().optional()
+});
+var sonioxTranslationStatusSchema = zod11.enum(["original", "translation", "none"]);
 var sonioxTokenSchema = zod11.object({
-  text: zod11.string().describe("Token text content (subword, word, or space)"),
-  start_ms: zod11.number().optional().describe("Start time of the token in milliseconds"),
-  end_ms: zod11.number().optional().describe("End time of the token in milliseconds"),
-  confidence: zod11.number().min(0).max(1).optional().describe("Confidence score between 0.0 and 1.0"),
-  is_final: zod11.boolean().describe("Whether this token is final (confirmed) or provisional"),
-  speaker: zod11.string().optional().describe("Speaker identifier (only present when speaker diarization is enabled)"),
-  language: zod11.string().optional().describe("Detected language code (only present when language identification is enabled)"),
-  source_language: zod11.string().optional().describe("Original language code for translated tokens"),
-  translation_status: sonioxTranslationStatusSchema.optional().describe("Translation status: 'none', 'original', or 'translation'")
+  text: zod11.string(),
+  start_ms: zod11.number().optional(),
+  end_ms: zod11.number().optional(),
+  confidence: zod11.number(),
+  is_final: zod11.boolean(),
+  speaker: zod11.string().optional(),
+  translation_status: sonioxTranslationStatusSchema.optional(),
+  language: zod11.string().optional(),
+  source_language: zod11.string().optional()
 });
 var sonioxStreamingResponseSchema = zod11.object({
-  text: zod11.string().optional().describe("Complete transcribed text"),
-  tokens: zod11.array(sonioxTokenSchema).describe("List of recognized tokens"),
-  final_audio_proc_ms: zod11.number().optional().describe("Milliseconds of audio processed into final tokens"),
-  total_audio_proc_ms: zod11.number().optional().describe("Milliseconds of audio processed (final + non-final)"),
-  finished: zod11.boolean().optional().describe("Whether the transcription is complete"),
-  error: zod11.string().optional().describe("Error message if an error occurred"),
-  error_code: zod11.number().optional().describe("Error code if an error occurred")
+  text: zod11.string(),
+  tokens: zod11.array(sonioxTokenSchema),
+  final_audio_proc_ms: zod11.number(),
+  total_audio_proc_ms: zod11.number(),
+  finished: zod11.boolean().optional(),
+  error_code: zod11.number().optional(),
+  error_message: zod11.string().optional()
 });
 var sonioxRecorderStateSchema = zod11.enum([
   "Init",
@@ -37176,8 +36786,8 @@ var BatchOnlyProviders = AllProviders.filter(
 );
 
 // src/generated/deepgram/schema/index.ts
-var schema_exports4 = {};
-__export(schema_exports4, {
+var schema_exports5 = {};
+__export(schema_exports5, {
   V1ListenPostParametersCallbackMethod: () => V1ListenPostParametersCallbackMethod,
   V1ListenPostParametersCustomIntentMode: () => V1ListenPostParametersCustomIntentMode,
   V1ListenPostParametersCustomTopicMode: () => V1ListenPostParametersCustomTopicMode,
@@ -37432,8 +37042,8 @@ var V1SpeakPostParametersSampleRate = {
 };
 
 // src/generated/openai/schema/index.ts
-var schema_exports5 = {};
-__export(schema_exports5, {
+var schema_exports6 = {};
+__export(schema_exports6, {
   AudioResponseFormat: () => AudioResponseFormat,
   CreateSpeechRequestResponseFormat: () => CreateSpeechRequestResponseFormat,
   CreateSpeechRequestStreamFormat: () => CreateSpeechRequestStreamFormat,
@@ -37727,6 +37337,16 @@ var ToolChoiceOptions = {
   required: "required"
 };
 
+// src/generated/openai/schema/transcriptionDiarizedSegmentType.ts
+var TranscriptionDiarizedSegmentType = {
+  transcripttextsegment: "transcript.text.segment"
+};
+
+// src/generated/openai/schema/transcriptionInclude.ts
+var TranscriptionInclude = {
+  logprobs: "logprobs"
+};
+
 // src/generated/openai/schema/transcriptTextDeltaEventType.ts
 var TranscriptTextDeltaEventType = {
   transcripttextdelta: "transcript.text.delta"
@@ -37752,16 +37372,6 @@ var TranscriptTextUsageTokensType = {
   tokens: "tokens"
 };
 
-// src/generated/openai/schema/transcriptionDiarizedSegmentType.ts
-var TranscriptionDiarizedSegmentType = {
-  transcripttextsegment: "transcript.text.segment"
-};
-
-// src/generated/openai/schema/transcriptionInclude.ts
-var TranscriptionInclude = {
-  logprobs: "logprobs"
-};
-
 // src/generated/openai/schema/vadConfigType.ts
 var VadConfigType = {
   server_vad: "server_vad"
@@ -37773,8 +37383,8 @@ var VoiceResourceObject = {
 };
 
 // src/generated/speechmatics/schema/index.ts
-var schema_exports6 = {};
-__export(schema_exports6, {
+var schema_exports7 = {};
+__export(schema_exports7, {
   AutoChaptersResultErrorType: () => AutoChaptersResultErrorType,
   ErrorResponseError: () => ErrorResponseError,
   GetJobsJobidAlignmentTags: () => GetJobsJobidAlignmentTags,
@@ -37963,32 +37573,6 @@ var WrittenFormRecognitionResultType = {
   word: "word"
 };
 
-// src/generated/soniox/schema/index.ts
-var schema_exports7 = {};
-__export(schema_exports7, {
-  TemporaryApiKeyUsageType: () => TemporaryApiKeyUsageType,
-  TranscriptionMode: () => TranscriptionMode,
-  TranscriptionStatus: () => TranscriptionStatus,
-  TranslationConfigType: () => TranslationConfigType
-});
-
-// src/generated/soniox/schema/temporaryApiKeyUsageType.ts
-var TemporaryApiKeyUsageType = {
-  transcribe_websocket: "transcribe_websocket"
-};
-
-// src/generated/soniox/schema/transcriptionMode.ts
-var TranscriptionMode = {
-  real_time: "real_time",
-  async: "async"
-};
-
-// src/generated/soniox/schema/translationConfigType.ts
-var TranslationConfigType = {
-  one_way: "one_way",
-  two_way: "two_way"
-};
-
 // src/generated/elevenlabs/schema/index.ts
 var schema_exports8 = {};
 __export(schema_exports8, {
@@ -38143,8 +37727,8 @@ var getJobsQueryParams = zod12.object({
 var getJobsResponseJobsItemDurationMin = 0;
 var getJobsResponseJobsItemConfigTranscriptionConfigPunctuationOverridesSensitivityMin = 0;
 var getJobsResponseJobsItemConfigTranscriptionConfigPunctuationOverridesSensitivityMax = 1;
-var getJobsResponseJobsItemConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = new RegExp("^(.|all)$");
-var getJobsResponseJobsItemConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = new RegExp("^[A-Za-z0-9._]+$");
+var getJobsResponseJobsItemConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = /^(.|all)$/;
+var getJobsResponseJobsItemConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = /^[A-Za-z0-9._]+$/;
 var getJobsResponseJobsItemConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMin = 0;
 var getJobsResponseJobsItemConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMax = 1;
 var getJobsResponseJobsItemConfigTranslationConfigTargetLanguagesMax = 5;
@@ -38342,8 +37926,8 @@ var getJobsJobidParams = zod12.object({
 var getJobsJobidResponseJobDurationMin = 0;
 var getJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesSensitivityMin = 0;
 var getJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesSensitivityMax = 1;
-var getJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = new RegExp("^(.|all)$");
-var getJobsJobidResponseJobConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = new RegExp("^[A-Za-z0-9._]+$");
+var getJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = /^(.|all)$/;
+var getJobsJobidResponseJobConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = /^[A-Za-z0-9._]+$/;
 var getJobsJobidResponseJobConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMin = 0;
 var getJobsJobidResponseJobConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMax = 1;
 var getJobsJobidResponseJobConfigTranslationConfigTargetLanguagesMax = 5;
@@ -38540,8 +38124,8 @@ var deleteJobsJobidQueryParams = zod12.object({
 var deleteJobsJobidResponseJobDurationMin = 0;
 var deleteJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesSensitivityMin = 0;
 var deleteJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesSensitivityMax = 1;
-var deleteJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = new RegExp("^(.|all)$");
-var deleteJobsJobidResponseJobConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = new RegExp("^[A-Za-z0-9._]+$");
+var deleteJobsJobidResponseJobConfigTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = /^(.|all)$/;
+var deleteJobsJobidResponseJobConfigTranscriptionConfigChannelDiarizationLabelsItemRegExp = /^[A-Za-z0-9._]+$/;
 var deleteJobsJobidResponseJobConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMin = 0;
 var deleteJobsJobidResponseJobConfigTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMax = 1;
 var deleteJobsJobidResponseJobConfigTranslationConfigTargetLanguagesMax = 5;
@@ -38744,8 +38328,8 @@ var getJobsJobidTranscriptQueryParams = zod12.object({
 var getJobsJobidTranscriptResponseJobDurationMin = 0;
 var getJobsJobidTranscriptResponseMetadataTranscriptionConfigPunctuationOverridesSensitivityMin = 0;
 var getJobsJobidTranscriptResponseMetadataTranscriptionConfigPunctuationOverridesSensitivityMax = 1;
-var getJobsJobidTranscriptResponseMetadataTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = new RegExp("^(.|all)$");
-var getJobsJobidTranscriptResponseMetadataTranscriptionConfigChannelDiarizationLabelsItemRegExp = new RegExp("^[A-Za-z0-9._]+$");
+var getJobsJobidTranscriptResponseMetadataTranscriptionConfigPunctuationOverridesPermittedMarksItemRegExp = /^(.|all)$/;
+var getJobsJobidTranscriptResponseMetadataTranscriptionConfigChannelDiarizationLabelsItemRegExp = /^[A-Za-z0-9._]+$/;
 var getJobsJobidTranscriptResponseMetadataTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMin = 0;
 var getJobsJobidTranscriptResponseMetadataTranscriptionConfigSpeakerDiarizationConfigSpeakerSensitivityMax = 1;
 var getJobsJobidTranscriptResponseResultsItemVolumeMin = 0;
@@ -39692,7 +39276,7 @@ export {
   DeepgramTTSSampleRate,
   DeepgramTopicMode,
   DeepgramTranscriptionSchema,
-  schema_exports4 as DeepgramTypes,
+  schema_exports5 as DeepgramTypes,
   deepgramAPI_zod_exports as DeepgramZodSchemas,
   ElevenLabsAdapter,
   ElevenLabsCapabilities,
@@ -39729,7 +39313,7 @@ export {
   OpenAIResponseFormat,
   streaming_types_exports as OpenAIStreamingTypes,
   OpenAITranscriptionSchema,
-  schema_exports5 as OpenAITypes,
+  schema_exports6 as OpenAITypes,
   OpenAIWhisperAdapter,
   openAIAudioRealtimeAPI_zod_exports as OpenAIZodSchemas,
   ProfanityFilterMode,
@@ -39758,7 +39342,7 @@ export {
   SonioxStreamingUpdateSchema,
   streaming_types_zod_exports as SonioxStreamingZodSchemas,
   SonioxTranscriptionSchema,
-  schema_exports7 as SonioxTypes,
+  schema_exports4 as SonioxTypes,
   SpeakV1ContainerParameter,
   SpeakV1EncodingParameter,
   SpeakV1SampleRateParameter,
@@ -39773,7 +39357,7 @@ export {
   SpeechmaticsStreamingSchema,
   SpeechmaticsStreamingUpdateSchema,
   SpeechmaticsTranscriptionSchema,
-  schema_exports6 as SpeechmaticsTypes,
+  schema_exports7 as SpeechmaticsTypes,
   speechmaticsASRRESTAPI_zod_exports as SpeechmaticsZodSchemas,
   StreamingProviders,
   StreamingSupportedBitDepthEnum,