@ai-sdk/assemblyai 3.0.0-beta.8 → 3.0.0-canary.33

package/package.json

@@ -1,10 +1,10 @@
 {
   "name": "@ai-sdk/assemblyai",
-  "version": "3.0.0-beta.8",
+  "version": "3.0.0-canary.33",
+  "type": "module",
   "license": "Apache-2.0",
   "sideEffects": false,
   "main": "./dist/index.js",
-  "module": "./dist/index.mjs",
   "types": "./dist/index.d.ts",
   "files": [
     "dist/**/*",
@@ -24,20 +24,20 @@
     "./package.json": "./package.json",
     ".": {
       "types": "./dist/index.d.ts",
-      "import": "./dist/index.mjs",
-      "require": "./dist/index.js"
+      "import": "./dist/index.js",
+      "default": "./dist/index.js"
     }
   },
   "dependencies": {
-    "@ai-sdk/provider": "4.0.0-beta.5",
-    "@ai-sdk/provider-utils": "5.0.0-beta.7"
+    "@ai-sdk/provider": "4.0.0-canary.15",
+    "@ai-sdk/provider-utils": "5.0.0-canary.31"
   },
   "devDependencies": {
     "@types/node": "20.17.24",
     "tsup": "^8",
     "typescript": "5.6.3",
     "zod": "3.25.76",
-    "@ai-sdk/test-server": "2.0.0-beta.0",
+    "@ai-sdk/test-server": "2.0.0-canary.4",
     "@vercel/ai-tsconfig": "0.0.0"
   },
   "peerDependencies": {
@@ -47,12 +47,14 @@
     "node": ">=18"
   },
   "publishConfig": {
-    "access": "public"
+    "access": "public",
+    "provenance": true
   },
   "homepage": "https://ai-sdk.dev/docs",
   "repository": {
     "type": "git",
-    "url": "git+https://github.com/vercel/ai.git"
+    "url": "https://github.com/vercel/ai",
+    "directory": "packages/assemblyai"
   },
   "bugs": {
     "url": "https://github.com/vercel/ai/issues"

package/src/assemblyai-config.ts

@@ -1,9 +1,9 @@
-import { FetchFunction } from '@ai-sdk/provider-utils';
+import type { FetchFunction } from '@ai-sdk/provider-utils';
 
 export type AssemblyAIConfig = {
   provider: string;
   url: (options: { modelId: string; path: string }) => string;
-  headers: () => Record<string, string | undefined>;
+  headers?: () => Record<string, string | undefined>;
   fetch?: FetchFunction;
   generateId?: () => string;
 };

package/src/assemblyai-provider.ts

@@ -1,15 +1,15 @@
 import {
-  TranscriptionModelV4,
-  ProviderV4,
   NoSuchModelError,
+  type TranscriptionModelV4,
+  type ProviderV4,
 } from '@ai-sdk/provider';
 import {
-  FetchFunction,
   loadApiKey,
   withUserAgentSuffix,
+  type FetchFunction,
 } from '@ai-sdk/provider-utils';
 import { AssemblyAITranscriptionModel } from './assemblyai-transcription-model';
-import { AssemblyAITranscriptionModelId } from './assemblyai-transcription-settings';
+import type { AssemblyAITranscriptionModelId } from './assemblyai-transcription-settings';
 import { VERSION } from './version';
 
 export interface AssemblyAIProvider extends ProviderV4 {

package/src/assemblyai-transcription-model-options.ts

@@ -0,0 +1,153 @@
+import { z } from 'zod/v4';
+
+// https://www.assemblyai.com/docs/api-reference/transcripts/submit
+export const assemblyaiTranscriptionModelOptionsSchema = z.object({
+  /**
+   * End time of the audio in milliseconds.
+   */
+  audioEndAt: z.number().int().nullish(),
+  /**
+   * Start time of the audio in milliseconds.
+   */
+  audioStartFrom: z.number().int().nullish(),
+  /**
+   * Whether to automatically generate chapters for the transcription.
+   */
+  autoChapters: z.boolean().nullish(),
+  /**
+   * Whether to automatically generate highlights for the transcription.
+   */
+  autoHighlights: z.boolean().nullish(),
+  /**
+   * Boost parameter for the transcription.
+   * Allowed values: 'low', 'default', 'high'.
+   */
+  boostParam: z.string().nullish(),
+  /**
+   * Whether to enable content safety filtering.
+   */
+  contentSafety: z.boolean().nullish(),
+  /**
+   * Confidence threshold for content safety filtering (25-100).
+   */
+  contentSafetyConfidence: z.number().int().min(25).max(100).nullish(),
+  /**
+   * Custom spelling rules for the transcription.
+   */
+  customSpelling: z
+    .array(
+      z.object({
+        from: z.array(z.string()),
+        to: z.string(),
+      }),
+    )
+    .nullish(),
+  /**
+   * Whether to include filler words (um, uh, etc.) in the transcription.
+   */
+  disfluencies: z.boolean().nullish(),
+  /**
+   * Whether to enable entity detection.
+   */
+  entityDetection: z.boolean().nullish(),
+  /**
+   * Whether to filter profanity from the transcription.
+   */
+  filterProfanity: z.boolean().nullish(),
+  /**
+   * Whether to format text with punctuation and capitalization.
+   */
+  formatText: z.boolean().nullish(),
+  /**
+   * Whether to enable IAB categories detection.
+   */
+  iabCategories: z.boolean().nullish(),
+  /**
+   * Language code for the transcription.
+   */
+  languageCode: z.union([z.literal('en'), z.string()]).nullish(),
+  /**
+   * Confidence threshold for language detection.
+   */
+  languageConfidenceThreshold: z.number().nullish(),
+  /**
+   * Whether to enable language detection.
+   */
+  languageDetection: z.boolean().nullish(),
+  /**
+   * Whether to process audio as multichannel.
+   */
+  multichannel: z.boolean().nullish(),
+  /**
+   * Whether to add punctuation to the transcription.
+   */
+  punctuate: z.boolean().nullish(),
+  /**
+   * Whether to redact personally identifiable information (PII).
+   */
+  redactPii: z.boolean().nullish(),
+  /**
+   * Whether to redact PII in the audio file.
+   */
+  redactPiiAudio: z.boolean().nullish(),
+  /**
+   * Audio format for PII redaction.
+   */
+  redactPiiAudioQuality: z.string().nullish(),
+  /**
+   * List of PII types to redact.
+   */
+  redactPiiPolicies: z.array(z.string()).nullish(),
+  /**
+   * Substitution method for redacted PII.
+   */
+  redactPiiSub: z.string().nullish(),
+  /**
+   * Whether to enable sentiment analysis.
+   */
+  sentimentAnalysis: z.boolean().nullish(),
+  /**
+   * Whether to identify different speakers in the audio.
+   */
+  speakerLabels: z.boolean().nullish(),
+  /**
+   * Number of speakers expected in the audio.
+   */
+  speakersExpected: z.number().int().nullish(),
+  /**
+   * Threshold for speech detection (0-1).
+   */
+  speechThreshold: z.number().min(0).max(1).nullish(),
+  /**
+   * Whether to generate a summary of the transcription.
+   */
+  summarization: z.boolean().nullish(),
+  /**
+   * Model to use for summarization.
+   */
+  summaryModel: z.string().nullish(),
+  /**
+   * Type of summary to generate.
+   */
+  summaryType: z.string().nullish(),
+  /**
+   * Name of the authentication header for webhook requests.
+   */
+  webhookAuthHeaderName: z.string().nullish(),
+  /**
+   * Value of the authentication header for webhook requests.
+   */
+  webhookAuthHeaderValue: z.string().nullish(),
+  /**
+   * URL to send webhook notifications to.
+   */
+  webhookUrl: z.string().nullish(),
+  /**
+   * List of words to boost recognition for.
+   */
+  wordBoost: z.array(z.string()).nullish(),
+});
+
+export type AssemblyAITranscriptionModelOptions = z.infer<
+  typeof assemblyaiTranscriptionModelOptionsSchema
+>;

package/src/assemblyai-transcription-model.ts

@@ -1,4 +1,4 @@
-import { TranscriptionModelV4, SharedV4Warning } from '@ai-sdk/provider';
+import type { TranscriptionModelV4, SharedV4Warning } from '@ai-sdk/provider';
 import {
   combineHeaders,
   createJsonResponseHandler,
@@ -6,164 +6,16 @@ import {
   parseProviderOptions,
   postJsonToApi,
   postToApi,
+  serializeModelOptions,
+  WORKFLOW_SERIALIZE,
+  WORKFLOW_DESERIALIZE,
 } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
-import { AssemblyAIConfig } from './assemblyai-config';
+import type { AssemblyAIConfig } from './assemblyai-config';
 import { assemblyaiFailedResponseHandler } from './assemblyai-error';
-import { AssemblyAITranscriptionModelId } from './assemblyai-transcription-settings';
-import { AssemblyAITranscriptionAPITypes } from './assemblyai-api-types';
-
-// https://www.assemblyai.com/docs/api-reference/transcripts/submit
-const assemblyaiTranscriptionModelOptionsSchema = z.object({
-  /**
-   * End time of the audio in milliseconds.
-   */
-  audioEndAt: z.number().int().nullish(),
-  /**
-   * Start time of the audio in milliseconds.
-   */
-  audioStartFrom: z.number().int().nullish(),
-  /**
-   * Whether to automatically generate chapters for the transcription.
-   */
-  autoChapters: z.boolean().nullish(),
-  /**
-   * Whether to automatically generate highlights for the transcription.
-   */
-  autoHighlights: z.boolean().nullish(),
-  /**
-   * Boost parameter for the transcription.
-   * Allowed values: 'low', 'default', 'high'.
-   */
-  boostParam: z.string().nullish(),
-  /**
-   * Whether to enable content safety filtering.
-   */
-  contentSafety: z.boolean().nullish(),
-  /**
-   * Confidence threshold for content safety filtering (25-100).
-   */
-  contentSafetyConfidence: z.number().int().min(25).max(100).nullish(),
-  /**
-   * Custom spelling rules for the transcription.
-   */
-  customSpelling: z
-    .array(
-      z.object({
-        from: z.array(z.string()),
-        to: z.string(),
-      }),
-    )
-    .nullish(),
-  /**
-   * Whether to include filler words (um, uh, etc.) in the transcription.
-   */
-  disfluencies: z.boolean().nullish(),
-  /**
-   * Whether to enable entity detection.
-   */
-  entityDetection: z.boolean().nullish(),
-  /**
-   * Whether to filter profanity from the transcription.
-   */
-  filterProfanity: z.boolean().nullish(),
-  /**
-   * Whether to format text with punctuation and capitalization.
-   */
-  formatText: z.boolean().nullish(),
-  /**
-   * Whether to enable IAB categories detection.
-   */
-  iabCategories: z.boolean().nullish(),
-  /**
-   * Language code for the transcription.
-   */
-  languageCode: z.union([z.literal('en'), z.string()]).nullish(),
-  /**
-   * Confidence threshold for language detection.
-   */
-  languageConfidenceThreshold: z.number().nullish(),
-  /**
-   * Whether to enable language detection.
-   */
-  languageDetection: z.boolean().nullish(),
-  /**
-   * Whether to process audio as multichannel.
-   */
-  multichannel: z.boolean().nullish(),
-  /**
-   * Whether to add punctuation to the transcription.
-   */
-  punctuate: z.boolean().nullish(),
-  /**
-   * Whether to redact personally identifiable information (PII).
-   */
-  redactPii: z.boolean().nullish(),
-  /**
-   * Whether to redact PII in the audio file.
-   */
-  redactPiiAudio: z.boolean().nullish(),
-  /**
-   * Audio format for PII redaction.
-   */
-  redactPiiAudioQuality: z.string().nullish(),
-  /**
-   * List of PII types to redact.
-   */
-  redactPiiPolicies: z.array(z.string()).nullish(),
-  /**
-   * Substitution method for redacted PII.
-   */
-  redactPiiSub: z.string().nullish(),
-  /**
-   * Whether to enable sentiment analysis.
-   */
-  sentimentAnalysis: z.boolean().nullish(),
-  /**
-   * Whether to identify different speakers in the audio.
-   */
-  speakerLabels: z.boolean().nullish(),
-  /**
-   * Number of speakers expected in the audio.
-   */
-  speakersExpected: z.number().int().nullish(),
-  /**
-   * Threshold for speech detection (0-1).
-   */
-  speechThreshold: z.number().min(0).max(1).nullish(),
-  /**
-   * Whether to generate a summary of the transcription.
-   */
-  summarization: z.boolean().nullish(),
-  /**
-   * Model to use for summarization.
-   */
-  summaryModel: z.string().nullish(),
-  /**
-   * Type of summary to generate.
-   */
-  summaryType: z.string().nullish(),
-  /**
-   * Name of the authentication header for webhook requests.
-   */
-  webhookAuthHeaderName: z.string().nullish(),
-  /**
-   * Value of the authentication header for webhook requests.
-   */
-  webhookAuthHeaderValue: z.string().nullish(),
-  /**
-   * URL to send webhook notifications to.
-   */
-  webhookUrl: z.string().nullish(),
-  /**
-   * List of words to boost recognition for.
-   */
-  wordBoost: z.array(z.string()).nullish(),
-});
-
-export type AssemblyAITranscriptionModelOptions = z.infer<
-  typeof assemblyaiTranscriptionModelOptionsSchema
->;
+import { assemblyaiTranscriptionModelOptionsSchema } from './assemblyai-transcription-model-options';
+import type { AssemblyAITranscriptionModelId } from './assemblyai-transcription-settings';
+import type { AssemblyAITranscriptionAPITypes } from './assemblyai-api-types';
 
 interface AssemblyAITranscriptionModelConfig extends AssemblyAIConfig {
   _internal?: {
@@ -183,6 +35,20 @@ export class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
     return this.config.provider;
   }
 
+  static [WORKFLOW_SERIALIZE](model: AssemblyAITranscriptionModel) {
+    return serializeModelOptions({
+      modelId: model.modelId,
+      config: model.config,
+    });
+  }
+
+  static [WORKFLOW_DESERIALIZE](options: {
+    modelId: AssemblyAITranscriptionModelId;
+    config: AssemblyAITranscriptionModelConfig;
+  }) {
+    return new AssemblyAITranscriptionModel(options.modelId, options.config);
+  }
+
   constructor(
     readonly modelId: AssemblyAITranscriptionModelId,
     private readonly config: AssemblyAITranscriptionModelConfig,
@@ -201,7 +67,7 @@ export class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
     });
 
     const body: Omit<AssemblyAITranscriptionAPITypes, 'audio_url'> = {
-      speech_model: this.modelId,
+      speech_model: this.modelId as 'best' | 'nano',
     };
 
     // Add provider-specific options
@@ -289,7 +155,7 @@ export class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
         {
           method: 'GET',
           headers: combineHeaders(
-            this.config.headers(),
+            this.config.headers?.(),
             headers,
           ) as HeadersInit,
           signal: abortSignal,
@@ -340,7 +206,7 @@ export class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
       }),
       headers: {
         'Content-Type': 'application/octet-stream',
-        ...combineHeaders(this.config.headers(), options.headers),
+        ...combineHeaders(this.config.headers?.(), options.headers),
       },
       body: {
         content: options.audio,
@@ -361,7 +227,7 @@ export class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
         path: '/v2/transcript',
         modelId: this.modelId,
       }),
-      headers: combineHeaders(this.config.headers(), options.headers),
+      headers: combineHeaders(this.config.headers?.(), options.headers),
       body: {
         ...body,
         audio_url: uploadResponse.upload_url,

package/src/assemblyai-transcription-settings.ts

@@ -1 +1 @@
-export type AssemblyAITranscriptionModelId = 'best' | 'nano';
+export type AssemblyAITranscriptionModelId = 'best' | 'nano' | (string & {});

package/src/index.ts

@@ -3,5 +3,5 @@ export type {
   AssemblyAIProvider,
   AssemblyAIProviderSettings,
 } from './assemblyai-provider';
-export type { AssemblyAITranscriptionModelOptions } from './assemblyai-transcription-model';
+export type { AssemblyAITranscriptionModelOptions } from './assemblyai-transcription-model-options';
 export { VERSION } from './version';

package/dist/index.d.mts

@@ -1,123 +0,0 @@
-import { TranscriptionModelV4, ProviderV4 } from '@ai-sdk/provider';
-import { FetchFunction } from '@ai-sdk/provider-utils';
-import { z } from 'zod/v4';
-
-type AssemblyAIConfig = {
-    provider: string;
-    url: (options: {
-        modelId: string;
-        path: string;
-    }) => string;
-    headers: () => Record<string, string | undefined>;
-    fetch?: FetchFunction;
-    generateId?: () => string;
-};
-
-type AssemblyAITranscriptionModelId = 'best' | 'nano';
-
-declare const assemblyaiTranscriptionModelOptionsSchema: z.ZodObject<{
-    audioEndAt: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    audioStartFrom: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    autoChapters: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    autoHighlights: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    boostParam: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    contentSafety: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    contentSafetyConfidence: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    customSpelling: z.ZodOptional<z.ZodNullable<z.ZodArray<z.ZodObject<{
-        from: z.ZodArray<z.ZodString>;
-        to: z.ZodString;
-    }, z.core.$strip>>>>;
-    disfluencies: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    entityDetection: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    filterProfanity: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    formatText: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    iabCategories: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    languageCode: z.ZodOptional<z.ZodNullable<z.ZodUnion<readonly [z.ZodLiteral<"en">, z.ZodString]>>>;
-    languageConfidenceThreshold: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    languageDetection: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    multichannel: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    punctuate: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    redactPii: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    redactPiiAudio: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    redactPiiAudioQuality: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    redactPiiPolicies: z.ZodOptional<z.ZodNullable<z.ZodArray<z.ZodString>>>;
-    redactPiiSub: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    sentimentAnalysis: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    speakerLabels: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    speakersExpected: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    speechThreshold: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    summarization: z.ZodOptional<z.ZodNullable<z.ZodBoolean>>;
-    summaryModel: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    summaryType: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    webhookAuthHeaderName: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    webhookAuthHeaderValue: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    webhookUrl: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    wordBoost: z.ZodOptional<z.ZodNullable<z.ZodArray<z.ZodString>>>;
-}, z.core.$strip>;
-type AssemblyAITranscriptionModelOptions = z.infer<typeof assemblyaiTranscriptionModelOptionsSchema>;
-interface AssemblyAITranscriptionModelConfig extends AssemblyAIConfig {
-    _internal?: {
-        currentDate?: () => Date;
-    };
-    /**
-     * The polling interval for checking transcript status in milliseconds.
-     */
-    pollingInterval?: number;
-}
-declare class AssemblyAITranscriptionModel implements TranscriptionModelV4 {
-    readonly modelId: AssemblyAITranscriptionModelId;
-    private readonly config;
-    readonly specificationVersion = "v4";
-    private readonly POLLING_INTERVAL_MS;
-    get provider(): string;
-    constructor(modelId: AssemblyAITranscriptionModelId, config: AssemblyAITranscriptionModelConfig);
-    private getArgs;
-    /**
-     * Polls the given transcript until we have a status other than `processing` or `queued`.
-     *
-     * @see https://www.assemblyai.com/docs/getting-started/transcribe-an-audio-file#step-33
-     */
-    private waitForCompletion;
-    doGenerate(options: Parameters<TranscriptionModelV4['doGenerate']>[0]): Promise<Awaited<ReturnType<TranscriptionModelV4['doGenerate']>>>;
-}
-
-interface AssemblyAIProvider extends ProviderV4 {
-    (modelId: 'best', settings?: {}): {
-        transcription: AssemblyAITranscriptionModel;
-    };
-    /**
-     * Creates a model for transcription.
-     */
-    transcription(modelId: AssemblyAITranscriptionModelId): TranscriptionModelV4;
-    /**
-     * @deprecated Use `embeddingModel` instead.
-     */
-    textEmbeddingModel(modelId: string): never;
-}
-interface AssemblyAIProviderSettings {
-    /**
-     * 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 AssemblyAI provider instance.
- */
-declare function createAssemblyAI(options?: AssemblyAIProviderSettings): AssemblyAIProvider;
-/**
- * Default AssemblyAI provider instance.
- */
-declare const assemblyai: AssemblyAIProvider;
-
-declare const VERSION: string;
-
-export { type AssemblyAIProvider, type AssemblyAIProviderSettings, type AssemblyAITranscriptionModelOptions, VERSION, assemblyai, createAssemblyAI };