@ai-sdk/elevenlabs 3.0.0-beta.3 → 3.0.0-beta.30

package/src/elevenlabs-provider.ts

@@ -1,21 +1,21 @@
 import {
-  TranscriptionModelV3,
-  SpeechModelV3,
-  ProviderV3,
   NoSuchModelError,
+  type TranscriptionModelV4,
+  type SpeechModelV4,
+  type ProviderV4,
 } from '@ai-sdk/provider';
 import {
-  FetchFunction,
   loadApiKey,
   withUserAgentSuffix,
+  type FetchFunction,
 } from '@ai-sdk/provider-utils';
 import { ElevenLabsTranscriptionModel } from './elevenlabs-transcription-model';
-import { ElevenLabsTranscriptionModelId } from './elevenlabs-transcription-options';
+import type { ElevenLabsTranscriptionModelId } from './elevenlabs-transcription-options';
 import { ElevenLabsSpeechModel } from './elevenlabs-speech-model';
-import { ElevenLabsSpeechModelId } from './elevenlabs-speech-options';
+import type { ElevenLabsSpeechModelId } from './elevenlabs-speech-options';
 import { VERSION } from './version';
 
-export interface ElevenLabsProvider extends ProviderV3 {
+export interface ElevenLabsProvider extends ProviderV4 {
   (
     modelId: 'scribe_v1',
     settings?: {},
@@ -26,12 +26,12 @@ export interface ElevenLabsProvider extends ProviderV3 {
   /**
    * Creates a model for transcription.
    */
-  transcription(modelId: ElevenLabsTranscriptionModelId): TranscriptionModelV3;
+  transcription(modelId: ElevenLabsTranscriptionModelId): TranscriptionModelV4;
 
   /**
    * Creates a model for speech generation.
    */
-  speech(modelId: ElevenLabsSpeechModelId): SpeechModelV3;
+  speech(modelId: ElevenLabsSpeechModelId): SpeechModelV4;
 
   /**
    * @deprecated Use `embeddingModel` instead.
@@ -98,7 +98,7 @@ export function createElevenLabs(
     };
   };
 
-  provider.specificationVersion = 'v3' as const;
+  provider.specificationVersion = 'v4' as const;
   provider.transcription = createTranscriptionModel;
   provider.transcriptionModel = createTranscriptionModel;
   provider.speech = createSpeechModel;

package/src/elevenlabs-speech-model.ts

@@ -1,15 +1,18 @@
-import { SpeechModelV3, SharedV3Warning } from '@ai-sdk/provider';
+import type { SpeechModelV4, SharedV4Warning } from '@ai-sdk/provider';
 import {
   combineHeaders,
   createBinaryResponseHandler,
   parseProviderOptions,
   postJsonToApi,
+  serializeModelOptions,
+  WORKFLOW_SERIALIZE,
+  WORKFLOW_DESERIALIZE,
 } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
-import { ElevenLabsConfig } from './elevenlabs-config';
+import type { ElevenLabsConfig } from './elevenlabs-config';
 import { elevenlabsFailedResponseHandler } from './elevenlabs-error';
-import { ElevenLabsSpeechAPITypes } from './elevenlabs-speech-api-types';
-import {
+import type { ElevenLabsSpeechAPITypes } from './elevenlabs-speech-api-types';
+import type {
   ElevenLabsSpeechModelId,
   ElevenLabsSpeechVoiceId,
 } from './elevenlabs-speech-options';
@@ -54,13 +57,27 @@ interface ElevenLabsSpeechModelConfig extends ElevenLabsConfig {
   };
 }
 
-export class ElevenLabsSpeechModel implements SpeechModelV3 {
-  readonly specificationVersion = 'v3';
+export class ElevenLabsSpeechModel implements SpeechModelV4 {
+  readonly specificationVersion = 'v4';
 
   get provider(): string {
     return this.config.provider;
   }
 
+  static [WORKFLOW_SERIALIZE](model: ElevenLabsSpeechModel) {
+    return serializeModelOptions({
+      modelId: model.modelId,
+      config: model.config,
+    });
+  }
+
+  static [WORKFLOW_DESERIALIZE](options: {
+    modelId: ElevenLabsSpeechModelId;
+    config: ElevenLabsSpeechModelConfig;
+  }) {
+    return new ElevenLabsSpeechModel(options.modelId, options.config);
+  }
+
   constructor(
     readonly modelId: ElevenLabsSpeechModelId,
     private readonly config: ElevenLabsSpeechModelConfig,
@@ -74,8 +91,8 @@ export class ElevenLabsSpeechModel implements SpeechModelV3 {
     language,
     speed,
     providerOptions,
-  }: Parameters<SpeechModelV3['doGenerate']>[0]) {
-    const warnings: SharedV3Warning[] = [];
+  }: Parameters<SpeechModelV4['doGenerate']>[0]) {
+    const warnings: SharedV4Warning[] = [];
 
     // Parse provider options
     const elevenLabsOptions = await parseProviderOptions({
@@ -214,8 +231,8 @@ export class ElevenLabsSpeechModel implements SpeechModelV3 {
   }
 
   async doGenerate(
-    options: Parameters<SpeechModelV3['doGenerate']>[0],
-  ): Promise<Awaited<ReturnType<SpeechModelV3['doGenerate']>>> {
+    options: Parameters<SpeechModelV4['doGenerate']>[0],
+  ): Promise<Awaited<ReturnType<SpeechModelV4['doGenerate']>>> {
     const currentDate = this.config._internal?.currentDate?.() ?? new Date();
     const { requestBody, queryParams, warnings, voiceId } =
       await this.getArgs(options);
@@ -233,7 +250,7 @@ export class ElevenLabsSpeechModel implements SpeechModelV3 {
         const queryString = new URLSearchParams(queryParams).toString();
         return queryString ? `${baseUrl}?${queryString}` : baseUrl;
       })(),
-      headers: combineHeaders(this.config.headers(), options.headers),
+      headers: combineHeaders(this.config.headers?.(), options.headers),
       body: requestBody,
       failedResponseHandler: elevenlabsFailedResponseHandler,
       successfulResponseHandler: createBinaryResponseHandler(),

package/src/elevenlabs-transcription-model.ts

@@ -1,4 +1,4 @@
-import { TranscriptionModelV3, SharedV3Warning } from '@ai-sdk/provider';
+import type { TranscriptionModelV4, SharedV4Warning } from '@ai-sdk/provider';
 import {
   combineHeaders,
   convertBase64ToUint8Array,
@@ -6,12 +6,15 @@ import {
   mediaTypeToExtension,
   parseProviderOptions,
   postFormDataToApi,
+  serializeModelOptions,
+  WORKFLOW_SERIALIZE,
+  WORKFLOW_DESERIALIZE,
 } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
-import { ElevenLabsConfig } from './elevenlabs-config';
+import type { ElevenLabsConfig } from './elevenlabs-config';
 import { elevenlabsFailedResponseHandler } from './elevenlabs-error';
-import { ElevenLabsTranscriptionModelId } from './elevenlabs-transcription-options';
-import { ElevenLabsTranscriptionAPITypes } from './elevenlabs-api-types';
+import type { ElevenLabsTranscriptionModelId } from './elevenlabs-transcription-options';
+import type { ElevenLabsTranscriptionAPITypes } from './elevenlabs-api-types';
 
 // https://elevenlabs.io/docs/api-reference/speech-to-text/convert
 const elevenLabsTranscriptionModelOptionsSchema = z.object({
@@ -36,13 +39,27 @@ interface ElevenLabsTranscriptionModelConfig extends ElevenLabsConfig {
   };
 }
 
-export class ElevenLabsTranscriptionModel implements TranscriptionModelV3 {
-  readonly specificationVersion = 'v3';
+export class ElevenLabsTranscriptionModel implements TranscriptionModelV4 {
+  readonly specificationVersion = 'v4';
 
   get provider(): string {
     return this.config.provider;
   }
 
+  static [WORKFLOW_SERIALIZE](model: ElevenLabsTranscriptionModel) {
+    return serializeModelOptions({
+      modelId: model.modelId,
+      config: model.config,
+    });
+  }
+
+  static [WORKFLOW_DESERIALIZE](options: {
+    modelId: ElevenLabsTranscriptionModelId;
+    config: ElevenLabsTranscriptionModelConfig;
+  }) {
+    return new ElevenLabsTranscriptionModel(options.modelId, options.config);
+  }
+
   constructor(
     readonly modelId: ElevenLabsTranscriptionModelId,
     private readonly config: ElevenLabsTranscriptionModelConfig,
@@ -52,8 +69,8 @@ export class ElevenLabsTranscriptionModel implements TranscriptionModelV3 {
     audio,
     mediaType,
     providerOptions,
-  }: Parameters<TranscriptionModelV3['doGenerate']>[0]) {
-    const warnings: SharedV3Warning[] = [];
+  }: Parameters<TranscriptionModelV4['doGenerate']>[0]) {
+    const warnings: SharedV4Warning[] = [];
 
     // Parse provider options
     const elevenlabsOptions = await parseProviderOptions({
@@ -111,8 +128,8 @@ export class ElevenLabsTranscriptionModel implements TranscriptionModelV3 {
   }
 
   async doGenerate(
-    options: Parameters<TranscriptionModelV3['doGenerate']>[0],
-  ): Promise<Awaited<ReturnType<TranscriptionModelV3['doGenerate']>>> {
+    options: Parameters<TranscriptionModelV4['doGenerate']>[0],
+  ): Promise<Awaited<ReturnType<TranscriptionModelV4['doGenerate']>>> {
     const currentDate = this.config._internal?.currentDate?.() ?? new Date();
     const { formData, warnings } = await this.getArgs(options);
 
@@ -125,7 +142,7 @@ export class ElevenLabsTranscriptionModel implements TranscriptionModelV3 {
         path: '/v1/speech-to-text',
         modelId: this.modelId,
       }),
-      headers: combineHeaders(this.config.headers(), options.headers),
+      headers: combineHeaders(this.config.headers?.(), options.headers),
       formData,
       failedResponseHandler: elevenlabsFailedResponseHandler,
       successfulResponseHandler: createJsonResponseHandler(

package/dist/index.d.mts

@@ -1,122 +0,0 @@
-import { TranscriptionModelV3, ProviderV3, SpeechModelV3 } from '@ai-sdk/provider';
-import { FetchFunction } from '@ai-sdk/provider-utils';
-import { z } from 'zod/v4';
-
-type ElevenLabsConfig = {
-    provider: string;
-    url: (options: {
-        modelId: string;
-        path: string;
-    }) => string;
-    headers: () => Record<string, string | undefined>;
-    fetch?: FetchFunction;
-    generateId?: () => string;
-};
-
-type ElevenLabsTranscriptionModelId = 'scribe_v1' | 'scribe_v1_experimental' | (string & {});
-
-declare const elevenLabsTranscriptionModelOptionsSchema: z.ZodObject<{
-    languageCode: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    tagAudioEvents: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    numSpeakers: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    timestampsGranularity: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-        none: "none";
-        word: "word";
-        character: "character";
-    }>>>>;
-    diarize: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    fileFormat: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-        pcm_s16le_16: "pcm_s16le_16";
-        other: "other";
-    }>>>>;
-}, z.core.$strip>;
-type ElevenLabsTranscriptionModelOptions = z.infer<typeof elevenLabsTranscriptionModelOptionsSchema>;
-interface ElevenLabsTranscriptionModelConfig extends ElevenLabsConfig {
-    _internal?: {
-        currentDate?: () => Date;
-    };
-}
-declare class ElevenLabsTranscriptionModel implements TranscriptionModelV3 {
-    readonly modelId: ElevenLabsTranscriptionModelId;
-    private readonly config;
-    readonly specificationVersion = "v3";
-    get provider(): string;
-    constructor(modelId: ElevenLabsTranscriptionModelId, config: ElevenLabsTranscriptionModelConfig);
-    private getArgs;
-    doGenerate(options: Parameters<TranscriptionModelV3['doGenerate']>[0]): Promise<Awaited<ReturnType<TranscriptionModelV3['doGenerate']>>>;
-}
-
-type ElevenLabsSpeechModelId = 'eleven_v3' | 'eleven_multilingual_v2' | 'eleven_flash_v2_5' | 'eleven_flash_v2' | 'eleven_turbo_v2_5' | 'eleven_turbo_v2' | 'eleven_monolingual_v1' | 'eleven_multilingual_v1' | (string & {});
-type ElevenLabsSpeechVoiceId = string;
-
-interface ElevenLabsProvider extends ProviderV3 {
-    (modelId: 'scribe_v1', settings?: {}): {
-        transcription: ElevenLabsTranscriptionModel;
-    };
-    /**
-     * Creates a model for transcription.
-     */
-    transcription(modelId: ElevenLabsTranscriptionModelId): TranscriptionModelV3;
-    /**
-     * Creates a model for speech generation.
-     */
-    speech(modelId: ElevenLabsSpeechModelId): SpeechModelV3;
-    /**
-     * @deprecated Use `embeddingModel` instead.
-     */
-    textEmbeddingModel(modelId: string): never;
-}
-interface ElevenLabsProviderSettings {
-    /**
-     * API key for authenticating requests.
-     */
-    apiKey?: string;
-    /**
-     * Custom headers to include in the requests.
-     */
-    headers?: Record<string, string>;
-    /**
-     * Custom fetch implementation. You can use it as a middleware to intercept requests,
-     * or to provide a custom fetch implementation for e.g. testing.
-     */
-    fetch?: FetchFunction;
-}
-/**
- * Create an ElevenLabs provider instance.
- */
-declare function createElevenLabs(options?: ElevenLabsProviderSettings): ElevenLabsProvider;
-/**
- * Default ElevenLabs provider instance.
- */
-declare const elevenlabs: ElevenLabsProvider;
-
-declare const elevenLabsSpeechModelOptionsSchema: z.ZodObject<{
-    languageCode: z.ZodOptional<z.ZodString>;
-    voiceSettings: z.ZodOptional<z.ZodObject<{
-        stability: z.ZodOptional<z.ZodNumber>;
-        similarityBoost: z.ZodOptional<z.ZodNumber>;
-        style: z.ZodOptional<z.ZodNumber>;
-        useSpeakerBoost: z.ZodOptional<z.ZodBoolean>;
-    }, z.core.$strip>>;
-    pronunciationDictionaryLocators: z.ZodOptional<z.ZodArray<z.ZodObject<{
-        pronunciationDictionaryId: z.ZodString;
-        versionId: z.ZodOptional<z.ZodString>;
-    }, z.core.$strip>>>;
-    seed: z.ZodOptional<z.ZodNumber>;
-    previousText: z.ZodOptional<z.ZodString>;
-    nextText: z.ZodOptional<z.ZodString>;
-    previousRequestIds: z.ZodOptional<z.ZodArray<z.ZodString>>;
-    nextRequestIds: z.ZodOptional<z.ZodArray<z.ZodString>>;
-    applyTextNormalization: z.ZodOptional<z.ZodEnum<{
-        auto: "auto";
-        on: "on";
-        off: "off";
-    }>>;
-    applyLanguageTextNormalization: z.ZodOptional<z.ZodBoolean>;
-    enableLogging: z.ZodOptional<z.ZodBoolean>;
-}, z.core.$strip>;
-type ElevenLabsSpeechModelOptions = z.infer<typeof elevenLabsSpeechModelOptionsSchema>;
-
-declare const VERSION: string;
-
-export { type ElevenLabsProvider, type ElevenLabsProviderSettings, type ElevenLabsSpeechModelId, type ElevenLabsSpeechModelOptions, type ElevenLabsSpeechVoiceId, type ElevenLabsTranscriptionModelOptions, VERSION, createElevenLabs, elevenlabs };