@ai-sdk/assemblyai 0.0.0-1c33ba03-20260114162300 → 0.0.0-4115c213-20260122152721

package/CHANGELOG.md

@@ -1,11 +1,40 @@
 # @ai-sdk/assemblyai
 
-## 0.0.0-1c33ba03-20260114162300
+## 0.0.0-4115c213-20260122152721
 
 ### Patch Changes
 
-- Updated dependencies [261c011]
-  - @ai-sdk/provider-utils@0.0.0-1c33ba03-20260114162300
+- 4caafb2: chore: excluded tests from src folder in npm package
+- Updated dependencies [4caafb2]
+  - @ai-sdk/provider@0.0.0-4115c213-20260122152721
+  - @ai-sdk/provider-utils@0.0.0-4115c213-20260122152721
+
+## 2.0.10
+
+### Patch Changes
+
+- 2b8369d: chore: add docs to package dist
+
+## 2.0.9
+
+### Patch Changes
+
+- 8dc54db: chore: add src folders to package bundle
+
+## 2.0.8
+
+### Patch Changes
+
+- Updated dependencies [5c090e7]
+  - @ai-sdk/provider@3.0.4
+  - @ai-sdk/provider-utils@4.0.8
+
+## 2.0.7
+
+### Patch Changes
+
+- Updated dependencies [46f46e4]
+  - @ai-sdk/provider-utils@4.0.7
 
 ## 2.0.6
 

package/dist/index.js

@@ -401,7 +401,7 @@ var assemblyaiTranscriptionResponseSchema = import_v42.z.object({
 });
 
 // src/version.ts
-var VERSION = true ? "0.0.0-1c33ba03-20260114162300" : "0.0.0-test";
+var VERSION = true ? "0.0.0-4115c213-20260122152721" : "0.0.0-test";
 
 // src/assemblyai-provider.ts
 function createAssemblyAI(options = {}) {

package/dist/index.mjs

@@ -385,7 +385,7 @@ var assemblyaiTranscriptionResponseSchema = z2.object({
 });
 
 // src/version.ts
-var VERSION = true ? "0.0.0-1c33ba03-20260114162300" : "0.0.0-test";
+var VERSION = true ? "0.0.0-4115c213-20260122152721" : "0.0.0-test";
 
 // src/assemblyai-provider.ts
 function createAssemblyAI(options = {}) {

package/docs/100-assemblyai.mdx

@@ -0,0 +1,282 @@
+---
+title: AssemblyAI
+description: Learn how to use the AssemblyAI provider for the AI SDK.
+---
+
+# AssemblyAI Provider
+
+The [AssemblyAI](https://assemblyai.com/) provider contains language model support for the AssemblyAI transcription API.
+
+## Setup
+
+The AssemblyAI provider is available in the `@ai-sdk/assemblyai` module. You can install it with
+
+<Tabs items={['pnpm', 'npm', 'yarn', 'bun']}>
+  <Tab>
+    <Snippet text="pnpm add @ai-sdk/assemblyai" dark />
+  </Tab>
+  <Tab>
+    <Snippet text="npm install @ai-sdk/assemblyai" dark />
+  </Tab>
+  <Tab>
+    <Snippet text="yarn add @ai-sdk/assemblyai" dark />
+  </Tab>
+
+  <Tab>
+    <Snippet text="bun add @ai-sdk/assemblyai" dark />
+  </Tab>
+</Tabs>
+
+## Provider Instance
+
+You can import the default provider instance `assemblyai` from `@ai-sdk/assemblyai`:
+
+```ts
+import { assemblyai } from '@ai-sdk/assemblyai';
+```
+
+If you need a customized setup, you can import `createAssemblyAI` from `@ai-sdk/assemblyai` and create a provider instance with your settings:
+
+```ts
+import { createAssemblyAI } from '@ai-sdk/assemblyai';
+
+const assemblyai = createAssemblyAI({
+  // custom settings, e.g.
+  fetch: customFetch,
+});
+```
+
+You can use the following optional settings to customize the AssemblyAI provider instance:
+
+- **apiKey** _string_
+
+  API key that is being sent using the `Authorization` header.
+  It defaults to the `ASSEMBLYAI_API_KEY` environment variable.
+
+- **headers** _Record&lt;string,string&gt;_
+
+  Custom headers to include in the requests.
+
+- **fetch** _(input: RequestInfo, init?: RequestInit) => Promise&lt;Response&gt;_
+
+  Custom [fetch](https://developer.mozilla.org/en-US/docs/Web/API/fetch) implementation.
+  Defaults to the global `fetch` function.
+  You can use it as a middleware to intercept requests,
+  or to provide a custom fetch implementation for e.g. testing.
+
+## Transcription Models
+
+You can create models that call the [AssemblyAI transcription API](https://www.assemblyai.com/docs/getting-started/transcribe-an-audio-file/typescript)
+using the `.transcription()` factory method.
+
+The first argument is the model id e.g. `best`.
+
+```ts
+const model = assemblyai.transcription('best');
+```
+
+You can also pass additional provider-specific options using the `providerOptions` argument. For example, supplying the `contentSafety` option will enable content safety filtering.
+
+```ts highlight="6"
+import { experimental_transcribe as transcribe } from 'ai';
+import { assemblyai } from '@ai-sdk/assemblyai';
+import { readFile } from 'fs/promises';
+
+const result = await transcribe({
+  model: assemblyai.transcription('best'),
+  audio: await readFile('audio.mp3'),
+  providerOptions: { assemblyai: { contentSafety: true } },
+});
+```
+
+The following provider options are available:
+
+- **audioEndAt** _number_
+
+  End time of the audio in milliseconds.
+  Optional.
+
+- **audioStartFrom** _number_
+
+  Start time of the audio in milliseconds.
+  Optional.
+
+- **autoChapters** _boolean_
+
+  Whether to automatically generate chapters for the transcription.
+  Optional.
+
+- **autoHighlights** _boolean_
+
+  Whether to automatically generate highlights for the transcription.
+  Optional.
+
+- **boostParam** _enum_
+
+  Boost parameter for the transcription.
+  Allowed values: `'low'`, `'default'`, `'high'`.
+  Optional.
+
+- **contentSafety** _boolean_
+
+  Whether to enable content safety filtering.
+  Optional.
+
+- **contentSafetyConfidence** _number_
+
+  Confidence threshold for content safety filtering (25-100).
+  Optional.
+
+- **customSpelling** _array of objects_
+
+  Custom spelling rules for the transcription.
+  Each object has `from` (array of strings) and `to` (string) properties.
+  Optional.
+
+- **disfluencies** _boolean_
+
+  Whether to include disfluencies (um, uh, etc.) in the transcription.
+  Optional.
+
+- **entityDetection** _boolean_
+
+  Whether to detect entities in the transcription.
+  Optional.
+
+- **filterProfanity** _boolean_
+
+  Whether to filter profanity in the transcription.
+  Optional.
+
+- **formatText** _boolean_
+
+  Whether to format the text in the transcription.
+  Optional.
+
+- **iabCategories** _boolean_
+
+  Whether to include IAB categories in the transcription.
+  Optional.
+
+- **languageCode** _string_
+
+  Language code for the audio.
+  Supports numerous ISO-639-1 and ISO-639-3 language codes.
+  Optional.
+
+- **languageConfidenceThreshold** _number_
+
+  Confidence threshold for language detection.
+  Optional.
+
+- **languageDetection** _boolean_
+
+  Whether to enable language detection.
+  Optional.
+
+- **multichannel** _boolean_
+
+  Whether to process multiple audio channels separately.
+  Optional.
+
+- **punctuate** _boolean_
+
+  Whether to add punctuation to the transcription.
+  Optional.
+
+- **redactPii** _boolean_
+
+  Whether to redact personally identifiable information.
+  Optional.
+
+- **redactPiiAudio** _boolean_
+
+  Whether to redact PII in the audio file.
+  Optional.
+
+- **redactPiiAudioQuality** _enum_
+
+  Quality of the redacted audio file.
+  Allowed values: `'mp3'`, `'wav'`.
+  Optional.
+
+- **redactPiiPolicies** _array of enums_
+
+  Policies for PII redaction, specifying which types of information to redact.
+  Supports numerous types like `'person_name'`, `'phone_number'`, etc.
+  Optional.
+
+- **redactPiiSub** _enum_
+
+  Substitution method for redacted PII.
+  Allowed values: `'entity_name'`, `'hash'`.
+  Optional.
+
+- **sentimentAnalysis** _boolean_
+
+  Whether to perform sentiment analysis on the transcription.
+  Optional.
+
+- **speakerLabels** _boolean_
+
+  Whether to label different speakers in the transcription.
+  Optional.
+
+- **speakersExpected** _number_
+
+  Expected number of speakers in the audio.
+  Optional.
+
+- **speechThreshold** _number_
+
+  Threshold for speech detection (0-1).
+  Optional.
+
+- **summarization** _boolean_
+
+  Whether to generate a summary of the transcription.
+  Optional.
+
+- **summaryModel** _enum_
+
+  Model to use for summarization.
+  Allowed values: `'informative'`, `'conversational'`, `'catchy'`.
+  Optional.
+
+- **summaryType** _enum_
+
+  Type of summary to generate.
+  Allowed values: `'bullets'`, `'bullets_verbose'`, `'gist'`, `'headline'`, `'paragraph'`.
+  Optional.
+
+- **topics** _array of strings_
+
+  List of topics to detect in the transcription.
+  Optional.
+
+- **webhookAuthHeaderName** _string_
+
+  Name of the authentication header for webhook requests.
+  Optional.
+
+- **webhookAuthHeaderValue** _string_
+
+  Value of the authentication header for webhook requests.
+  Optional.
+
+- **webhookUrl** _string_
+
+  URL to send webhook notifications to.
+  Optional.
+
+- **wordBoost** _array of strings_
+
+  List of words to boost in the transcription.
+  Optional.
+
+### Model Capabilities
+
+| Model  | Transcription       | Duration            | Segments            | Language            |
+| ------ | ------------------- | ------------------- | ------------------- | ------------------- |
+| `best` | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
+| `nano` | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ai-sdk/assemblyai",
-  "version": "0.0.0-1c33ba03-20260114162300",
+  "version": "0.0.0-4115c213-20260122152721",
   "license": "Apache-2.0",
   "sideEffects": false,
   "main": "./dist/index.js",
@@ -8,9 +8,18 @@
   "types": "./dist/index.d.ts",
   "files": [
     "dist/**/*",
+    "docs/**/*",
+    "src",
+    "!src/**/*.test.ts",
+    "!src/**/*.test-d.ts",
+    "!src/**/__snapshots__",
+    "!src/**/__fixtures__",
     "CHANGELOG.md",
     "README.md"
   ],
+  "directories": {
+    "doc": "./docs"
+  },
   "exports": {
     "./package.json": "./package.json",
     ".": {
@@ -20,15 +29,15 @@
     }
   },
   "dependencies": {
-    "@ai-sdk/provider": "3.0.3",
-    "@ai-sdk/provider-utils": "0.0.0-1c33ba03-20260114162300"
+    "@ai-sdk/provider": "0.0.0-4115c213-20260122152721",
+    "@ai-sdk/provider-utils": "0.0.0-4115c213-20260122152721"
   },
   "devDependencies": {
     "@types/node": "20.17.24",
     "tsup": "^8",
     "typescript": "5.6.3",
     "zod": "3.25.76",
-    "@ai-sdk/test-server": "1.0.1",
+    "@ai-sdk/test-server": "0.0.0-4115c213-20260122152721",
     "@vercel/ai-tsconfig": "0.0.0"
   },
   "peerDependencies": {
@@ -54,7 +63,7 @@
   "scripts": {
     "build": "pnpm clean && tsup --tsconfig tsconfig.build.json",
     "build:watch": "pnpm clean && tsup --watch --tsconfig tsconfig.build.json",
-    "clean": "del-cli dist *.tsbuildinfo",
+    "clean": "del-cli dist docs *.tsbuildinfo",
     "lint": "eslint \"./**/*.ts*\"",
     "type-check": "tsc --build",
     "prettier-check": "prettier --check \"./**/*.ts*\"",

package/src/assemblyai-api-types.ts

@@ -0,0 +1,362 @@
+export type AssemblyAITranscriptionAPITypes = {
+  /**
+   * The URL of the audio or video file to transcribe.
+   */
+  audio_url: string;
+
+  /**
+   * The point in time, in milliseconds, to stop transcribing in your media file
+   */
+  audio_end_at?: number;
+
+  /**
+   * The point in time, in milliseconds, to begin transcribing in your media file
+   */
+  audio_start_from?: number;
+
+  /**
+   * Enable Auto Chapters, can be true or false
+   * @default false
+   */
+  auto_chapters?: boolean;
+
+  /**
+   * Enable Key Phrases, either true or false
+   * @default false
+   */
+  auto_highlights?: boolean;
+
+  /**
+   * How much to boost specified words
+   */
+  boost_param?: 'low' | 'default' | 'high';
+
+  /**
+   * Enable Content Moderation, can be true or false
+   * @default false
+   */
+  content_safety?: boolean;
+
+  /**
+   * The confidence threshold for the Content Moderation model. Values must be between 25 and 100.
+   * @default 50
+   */
+  content_safety_confidence?: number;
+
+  /**
+   * Customize how words are spelled and formatted using to and from values
+   */
+  custom_spelling?: Array<{
+    /**
+     * Words or phrases to replace
+     */
+    from: string[];
+    /**
+     * Word to replace with
+     */
+    to: string;
+  }>;
+
+  /**
+   * Transcribe Filler Words, like "umm", in your media file; can be true or false
+   * @default false
+   */
+  disfluencies?: boolean;
+
+  /**
+   * Enable Entity Detection, can be true or false
+   * @default false
+   */
+  entity_detection?: boolean;
+
+  /**
+   * Filter profanity from the transcribed text, can be true or false
+   * @default false
+   */
+  filter_profanity?: boolean;
+
+  /**
+   * Enable Text Formatting, can be true or false
+   * @default true
+   */
+  format_text?: boolean;
+
+  /**
+   * Enable Topic Detection, can be true or false
+   * @default false
+   */
+  iab_categories?: boolean;
+
+  /**
+   * The language of your audio file. Possible values are found in Supported Languages.
+   * @default 'en_us'
+   */
+  language_code?:
+    | 'en'
+    | 'en_au'
+    | 'en_uk'
+    | 'en_us'
+    | 'es'
+    | 'fr'
+    | 'de'
+    | 'it'
+    | 'pt'
+    | 'nl'
+    | 'af'
+    | 'sq'
+    | 'am'
+    | 'ar'
+    | 'hy'
+    | 'as'
+    | 'az'
+    | 'ba'
+    | 'eu'
+    | 'be'
+    | 'bn'
+    | 'bs'
+    | 'br'
+    | 'bg'
+    | 'my'
+    | 'ca'
+    | 'zh'
+    | 'hr'
+    | 'cs'
+    | 'da'
+    | 'et'
+    | 'fo'
+    | 'fi'
+    | 'gl'
+    | 'ka'
+    | 'el'
+    | 'gu'
+    | 'ht'
+    | 'ha'
+    | 'haw'
+    | 'he'
+    | 'hi'
+    | 'hu'
+    | 'is'
+    | 'id'
+    | 'ja'
+    | 'jw'
+    | 'kn'
+    | 'kk'
+    | 'km'
+    | 'ko'
+    | 'lo'
+    | 'la'
+    | 'lv'
+    | 'ln'
+    | 'lt'
+    | 'lb'
+    | 'mk'
+    | 'mg'
+    | 'ms'
+    | 'ml'
+    | 'mt'
+    | 'mi'
+    | 'mr'
+    | 'mn'
+    | 'ne'
+    | 'no'
+    | 'nn'
+    | 'oc'
+    | 'pa'
+    | 'ps'
+    | 'fa'
+    | 'pl'
+    | 'ro'
+    | 'ru'
+    | 'sa'
+    | 'sr'
+    | 'sn'
+    | 'sd'
+    | 'si'
+    | 'sk'
+    | 'sl'
+    | 'so'
+    | 'su'
+    | 'sw'
+    | 'sv'
+    | 'tl'
+    | 'tg'
+    | 'ta'
+    | 'tt'
+    | 'te'
+    | 'th'
+    | 'bo'
+    | 'tr'
+    | 'tk'
+    | 'uk'
+    | 'ur'
+    | 'uz'
+    | 'vi'
+    | 'cy'
+    | 'yi'
+    | 'yo';
+
+  /**
+   * The confidence threshold for the automatically detected language. An error will be returned if the language confidence is below this threshold.
+   * @default 0
+   */
+  language_confidence_threshold?: number;
+
+  /**
+   * Enable Automatic language detection, either true or false.
+   * @default false
+   */
+  language_detection?: boolean;
+
+  /**
+   * Enable Multichannel transcription, can be true or false.
+   * @default false
+   */
+  multichannel?: boolean;
+
+  /**
+   * Enable Automatic Punctuation, can be true or false
+   * @default true
+   */
+  punctuate?: boolean;
+
+  /**
+   * Redact PII from the transcribed text using the Redact PII model, can be true or false
+   * @default false
+   */
+  redact_pii?: boolean;
+
+  /**
+   * Generate a copy of the original media file with spoken PII "beeped" out, can be true or false.
+   * @default false
+   */
+  redact_pii_audio?: boolean;
+
+  /**
+   * Controls the filetype of the audio created by redact_pii_audio. Currently supports mp3 (default) and wav.
+   */
+  redact_pii_audio_quality?: 'mp3' | 'wav';
+
+  /**
+   * The list of PII Redaction policies to enable.
+   */
+  redact_pii_policies?: Array<
+    | 'account_number'
+    | 'banking_information'
+    | 'blood_type'
+    | 'credit_card_cvv'
+    | 'credit_card_expiration'
+    | 'credit_card_number'
+    | 'date'
+    | 'date_interval'
+    | 'date_of_birth'
+    | 'drivers_license'
+    | 'drug'
+    | 'duration'
+    | 'email_address'
+    | 'event'
+    | 'filename'
+    | 'gender_sexuality'
+    | 'healthcare_number'
+    | 'injury'
+    | 'ip_address'
+    | 'language'
+    | 'location'
+    | 'marital_status'
+    | 'medical_condition'
+    | 'medical_process'
+    | 'money_amount'
+    | 'nationality'
+    | 'number_sequence'
+    | 'occupation'
+    | 'organization'
+    | 'passport_number'
+    | 'password'
+    | 'person_age'
+    | 'person_name'
+    | 'phone_number'
+    | 'physical_attribute'
+    | 'political_affiliation'
+    | 'religion'
+    | 'statistics'
+    | 'time'
+    | 'url'
+    | 'us_social_security_number'
+    | 'username'
+    | 'vehicle_id'
+    | 'zodiac_sign'
+  >;
+
+  /**
+   * The replacement logic for detected PII, can be "entity_name" or "hash".
+   */
+  redact_pii_sub?: 'entity_name' | 'hash';
+
+  /**
+   * Enable Sentiment Analysis, can be true or false
+   * @default false
+   */
+  sentiment_analysis?: boolean;
+
+  /**
+   * Enable Speaker diarization, can be true or false
+   * @default false
+   */
+  speaker_labels?: boolean;
+
+  /**
+   * Tells the speaker label model how many speakers it should attempt to identify, up to 10.
+   */
+  speakers_expected?: number;
+
+  /**
+   * The speech model to use for the transcription.
+   */
+  speech_model?: 'best' | 'nano';
+
+  /**
+   * Reject audio files that contain less than this fraction of speech. Valid values are in the range [0, 1] inclusive.
+   */
+  speech_threshold?: number;
+
+  /**
+   * Enable Summarization, can be true or false
+   * @default false
+   */
+  summarization?: boolean;
+
+  /**
+   * The model to summarize the transcript
+   */
+  summary_model?: 'informative' | 'conversational' | 'catchy';
+
+  /**
+   * The type of summary
+   */
+  summary_type?:
+    | 'bullets'
+    | 'bullets_verbose'
+    | 'gist'
+    | 'headline'
+    | 'paragraph';
+
+  /**
+   * The header name to be sent with the transcript completed or failed webhook requests
+   */
+  webhook_auth_header_name?: string;
+
+  /**
+   * The header value to send back with the transcript completed or failed webhook requests for added security
+   */
+  webhook_auth_header_value?: string;
+
+  /**
+   * The URL to which we send webhook requests. We sends two different types of webhook requests.
+   * One request when a transcript is completed or failed, and one request when the redacted audio is ready if redact_pii_audio is enabled.
+   */
+  webhook_url?: string;
+
+  /**
+   * The list of custom vocabulary to boost transcription probability for
+   */
+  word_boost?: string[];
+};

package/src/assemblyai-config.ts

@@ -0,0 +1,9 @@
+import { FetchFunction } from '@ai-sdk/provider-utils';
+
+export type AssemblyAIConfig = {
+  provider: string;
+  url: (options: { modelId: string; path: string }) => string;
+  headers: () => Record<string, string | undefined>;
+  fetch?: FetchFunction;
+  generateId?: () => string;
+};

package/src/assemblyai-error.ts

@@ -0,0 +1,16 @@
+import { z } from 'zod/v4';
+import { createJsonErrorResponseHandler } from '@ai-sdk/provider-utils';
+
+export const assemblyaiErrorDataSchema = z.object({
+  error: z.object({
+    message: z.string(),
+    code: z.number(),
+  }),
+});
+
+export type AssemblyAIErrorData = z.infer<typeof assemblyaiErrorDataSchema>;
+
+export const assemblyaiFailedResponseHandler = createJsonErrorResponseHandler({
+  errorSchema: assemblyaiErrorDataSchema,
+  errorToMessage: data => data.error.message,
+});

package/src/assemblyai-provider.ts

@@ -0,0 +1,112 @@
+import {
+  TranscriptionModelV3,
+  ProviderV3,
+  NoSuchModelError,
+} from '@ai-sdk/provider';
+import {
+  FetchFunction,
+  loadApiKey,
+  withUserAgentSuffix,
+} from '@ai-sdk/provider-utils';
+import { AssemblyAITranscriptionModel } from './assemblyai-transcription-model';
+import { AssemblyAITranscriptionModelId } from './assemblyai-transcription-settings';
+import { VERSION } from './version';
+
+export interface AssemblyAIProvider extends ProviderV3 {
+  (
+    modelId: 'best',
+    settings?: {},
+  ): {
+    transcription: AssemblyAITranscriptionModel;
+  };
+
+  /**
+Creates a model for transcription.
+   */
+  transcription(modelId: AssemblyAITranscriptionModelId): TranscriptionModelV3;
+
+  /**
+   * @deprecated Use `embeddingModel` instead.
+   */
+  textEmbeddingModel(modelId: string): never;
+}
+
+export 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.
+ */
+export function createAssemblyAI(
+  options: AssemblyAIProviderSettings = {},
+): AssemblyAIProvider {
+  const getHeaders = () =>
+    withUserAgentSuffix(
+      {
+        authorization: loadApiKey({
+          apiKey: options.apiKey,
+          environmentVariableName: 'ASSEMBLYAI_API_KEY',
+          description: 'AssemblyAI',
+        }),
+        ...options.headers,
+      },
+      `ai-sdk/assemblyai/${VERSION}`,
+    );
+
+  const createTranscriptionModel = (modelId: AssemblyAITranscriptionModelId) =>
+    new AssemblyAITranscriptionModel(modelId, {
+      provider: `assemblyai.transcription`,
+      url: ({ path }) => `https://api.assemblyai.com${path}`,
+      headers: getHeaders,
+      fetch: options.fetch,
+    });
+
+  const provider = function (modelId: AssemblyAITranscriptionModelId) {
+    return {
+      transcription: createTranscriptionModel(modelId),
+    };
+  };
+
+  provider.specificationVersion = 'v3' as const;
+  provider.transcription = createTranscriptionModel;
+  provider.transcriptionModel = createTranscriptionModel;
+
+  provider.languageModel = () => {
+    throw new NoSuchModelError({
+      modelId: 'unknown',
+      modelType: 'languageModel',
+      message: 'AssemblyAI does not provide language models',
+    });
+  };
+
+  provider.embeddingModel = (modelId: string) => {
+    throw new NoSuchModelError({ modelId, modelType: 'embeddingModel' });
+  };
+  provider.textEmbeddingModel = provider.embeddingModel;
+
+  provider.imageModel = (modelId: string) => {
+    throw new NoSuchModelError({ modelId, modelType: 'imageModel' });
+  };
+
+  return provider as AssemblyAIProvider;
+}
+
+/**
+Default AssemblyAI provider instance.
+ */
+export const assemblyai = createAssemblyAI();

package/src/assemblyai-transcription-model.ts

@@ -0,0 +1,430 @@
+import { TranscriptionModelV3, SharedV3Warning } from '@ai-sdk/provider';
+import {
+  combineHeaders,
+  createJsonResponseHandler,
+  extractResponseHeaders,
+  parseProviderOptions,
+  postJsonToApi,
+  postToApi,
+} from '@ai-sdk/provider-utils';
+import { z } from 'zod/v4';
+import { 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 assemblyaiProviderOptionsSchema = 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 AssemblyAITranscriptionCallOptions = z.infer<
+  typeof assemblyaiProviderOptionsSchema
+>;
+
+interface AssemblyAITranscriptionModelConfig extends AssemblyAIConfig {
+  _internal?: {
+    currentDate?: () => Date;
+  };
+  /**
+   * The polling interval for checking transcript status in milliseconds.
+   */
+  pollingInterval?: number;
+}
+
+export class AssemblyAITranscriptionModel implements TranscriptionModelV3 {
+  readonly specificationVersion = 'v3';
+  private readonly POLLING_INTERVAL_MS = 3000;
+
+  get provider(): string {
+    return this.config.provider;
+  }
+
+  constructor(
+    readonly modelId: AssemblyAITranscriptionModelId,
+    private readonly config: AssemblyAITranscriptionModelConfig,
+  ) {}
+
+  private async getArgs({
+    providerOptions,
+  }: Parameters<TranscriptionModelV3['doGenerate']>[0]) {
+    const warnings: SharedV3Warning[] = [];
+
+    // Parse provider options
+    const assemblyaiOptions = await parseProviderOptions({
+      provider: 'assemblyai',
+      providerOptions,
+      schema: assemblyaiProviderOptionsSchema,
+    });
+
+    const body: Omit<AssemblyAITranscriptionAPITypes, 'audio_url'> = {
+      speech_model: this.modelId,
+    };
+
+    // Add provider-specific options
+    if (assemblyaiOptions) {
+      body.audio_end_at = assemblyaiOptions.audioEndAt ?? undefined;
+      body.audio_start_from = assemblyaiOptions.audioStartFrom ?? undefined;
+      body.auto_chapters = assemblyaiOptions.autoChapters ?? undefined;
+      body.auto_highlights = assemblyaiOptions.autoHighlights ?? undefined;
+      body.boost_param = (assemblyaiOptions.boostParam as never) ?? undefined;
+      body.content_safety = assemblyaiOptions.contentSafety ?? undefined;
+      body.content_safety_confidence =
+        assemblyaiOptions.contentSafetyConfidence ?? undefined;
+      body.custom_spelling =
+        (assemblyaiOptions.customSpelling as never) ?? undefined;
+      body.disfluencies = assemblyaiOptions.disfluencies ?? undefined;
+      body.entity_detection = assemblyaiOptions.entityDetection ?? undefined;
+      body.filter_profanity = assemblyaiOptions.filterProfanity ?? undefined;
+      body.format_text = assemblyaiOptions.formatText ?? undefined;
+      body.iab_categories = assemblyaiOptions.iabCategories ?? undefined;
+      body.language_code =
+        (assemblyaiOptions.languageCode as never) ?? undefined;
+      body.language_confidence_threshold =
+        assemblyaiOptions.languageConfidenceThreshold ?? undefined;
+      body.language_detection =
+        assemblyaiOptions.languageDetection ?? undefined;
+      body.multichannel = assemblyaiOptions.multichannel ?? undefined;
+      body.punctuate = assemblyaiOptions.punctuate ?? undefined;
+      body.redact_pii = assemblyaiOptions.redactPii ?? undefined;
+      body.redact_pii_audio = assemblyaiOptions.redactPiiAudio ?? undefined;
+      body.redact_pii_audio_quality =
+        (assemblyaiOptions.redactPiiAudioQuality as never) ?? undefined;
+      body.redact_pii_policies =
+        (assemblyaiOptions.redactPiiPolicies as never) ?? undefined;
+      body.redact_pii_sub =
+        (assemblyaiOptions.redactPiiSub as never) ?? undefined;
+      body.sentiment_analysis =
+        assemblyaiOptions.sentimentAnalysis ?? undefined;
+      body.speaker_labels = assemblyaiOptions.speakerLabels ?? undefined;
+      body.speakers_expected = assemblyaiOptions.speakersExpected ?? undefined;
+      body.speech_threshold = assemblyaiOptions.speechThreshold ?? undefined;
+      body.summarization = assemblyaiOptions.summarization ?? undefined;
+      body.summary_model =
+        (assemblyaiOptions.summaryModel as never) ?? undefined;
+      body.summary_type = (assemblyaiOptions.summaryType as never) ?? undefined;
+      body.webhook_auth_header_name =
+        assemblyaiOptions.webhookAuthHeaderName ?? undefined;
+      body.webhook_auth_header_value =
+        assemblyaiOptions.webhookAuthHeaderValue ?? undefined;
+      body.webhook_url = assemblyaiOptions.webhookUrl ?? undefined;
+      body.word_boost = assemblyaiOptions.wordBoost ?? undefined;
+    }
+
+    return {
+      body,
+      warnings,
+    };
+  }
+
+  /**
+   * 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 async waitForCompletion(
+    transcriptId: string,
+    headers: Record<string, string | undefined> | undefined,
+    abortSignal?: AbortSignal,
+  ): Promise<{
+    transcript: z.infer<typeof assemblyaiTranscriptionResponseSchema>;
+    responseHeaders: Record<string, string>;
+  }> {
+    const pollingInterval =
+      this.config.pollingInterval ?? this.POLLING_INTERVAL_MS;
+
+    while (true) {
+      if (abortSignal?.aborted) {
+        throw new Error('Transcription request was aborted');
+      }
+
+      const response = await fetch(
+        this.config.url({
+          path: `/v2/transcript/${transcriptId}`,
+          modelId: this.modelId,
+        }),
+        {
+          method: 'GET',
+          headers: combineHeaders(
+            this.config.headers(),
+            headers,
+          ) as HeadersInit,
+          signal: abortSignal,
+        },
+      );
+
+      if (!response.ok) {
+        throw await assemblyaiFailedResponseHandler({
+          response,
+          url: this.config.url({
+            path: `/v2/transcript/${transcriptId}`,
+            modelId: this.modelId,
+          }),
+          requestBodyValues: {},
+        });
+      }
+
+      const transcript = assemblyaiTranscriptionResponseSchema.parse(
+        await response.json(),
+      );
+
+      if (transcript.status === 'completed') {
+        return {
+          transcript,
+          responseHeaders: extractResponseHeaders(response),
+        };
+      }
+
+      if (transcript.status === 'error') {
+        throw new Error(
+          `Transcription failed: ${transcript.error ?? 'Unknown error'}`,
+        );
+      }
+
+      await new Promise(resolve => setTimeout(resolve, pollingInterval));
+    }
+  }
+
+  async doGenerate(
+    options: Parameters<TranscriptionModelV3['doGenerate']>[0],
+  ): Promise<Awaited<ReturnType<TranscriptionModelV3['doGenerate']>>> {
+    const currentDate = this.config._internal?.currentDate?.() ?? new Date();
+
+    const { value: uploadResponse } = await postToApi({
+      url: this.config.url({
+        path: '/v2/upload',
+        modelId: '',
+      }),
+      headers: {
+        'Content-Type': 'application/octet-stream',
+        ...combineHeaders(this.config.headers(), options.headers),
+      },
+      body: {
+        content: options.audio,
+        values: options.audio,
+      },
+      failedResponseHandler: assemblyaiFailedResponseHandler,
+      successfulResponseHandler: createJsonResponseHandler(
+        assemblyaiUploadResponseSchema,
+      ),
+      abortSignal: options.abortSignal,
+      fetch: this.config.fetch,
+    });
+
+    const { body, warnings } = await this.getArgs(options);
+
+    const { value: submitResponse } = await postJsonToApi({
+      url: this.config.url({
+        path: '/v2/transcript',
+        modelId: this.modelId,
+      }),
+      headers: combineHeaders(this.config.headers(), options.headers),
+      body: {
+        ...body,
+        audio_url: uploadResponse.upload_url,
+      },
+      failedResponseHandler: assemblyaiFailedResponseHandler,
+      successfulResponseHandler: createJsonResponseHandler(
+        assemblyaiSubmitResponseSchema,
+      ),
+      abortSignal: options.abortSignal,
+      fetch: this.config.fetch,
+    });
+
+    const { transcript, responseHeaders } = await this.waitForCompletion(
+      submitResponse.id,
+      options.headers,
+      options.abortSignal,
+    );
+
+    return {
+      text: transcript.text ?? '',
+      segments:
+        transcript.words?.map(word => ({
+          text: word.text,
+          startSecond: word.start,
+          endSecond: word.end,
+        })) ?? [],
+      language: transcript.language_code ?? undefined,
+      durationInSeconds:
+        transcript.audio_duration ?? transcript.words?.at(-1)?.end ?? undefined,
+      warnings,
+      response: {
+        timestamp: currentDate,
+        modelId: this.modelId,
+        headers: responseHeaders, // Headers from final GET request
+        body: transcript, // Raw response from final GET request
+      },
+    };
+  }
+}
+
+const assemblyaiUploadResponseSchema = z.object({
+  upload_url: z.string(),
+});
+
+const assemblyaiSubmitResponseSchema = z.object({
+  id: z.string(),
+  status: z.enum(['queued', 'processing', 'completed', 'error']),
+});
+
+const assemblyaiTranscriptionResponseSchema = z.object({
+  id: z.string(),
+  status: z.enum(['queued', 'processing', 'completed', 'error']),
+  text: z.string().nullish(),
+  language_code: z.string().nullish(),
+  words: z
+    .array(
+      z.object({
+        start: z.number(),
+        end: z.number(),
+        text: z.string(),
+      }),
+    )
+    .nullish(),
+  audio_duration: z.number().nullish(),
+  error: z.string().nullish(),
+});

package/src/assemblyai-transcription-settings.ts

@@ -0,0 +1 @@
+export type AssemblyAITranscriptionModelId = 'best' | 'nano';

package/src/index.ts

@@ -0,0 +1,6 @@
+export { createAssemblyAI, assemblyai } from './assemblyai-provider';
+export type {
+  AssemblyAIProvider,
+  AssemblyAIProviderSettings,
+} from './assemblyai-provider';
+export { VERSION } from './version';

package/src/version.ts

@@ -0,0 +1,6 @@
+// Version string of this package injected at build time.
+declare const __PACKAGE_VERSION__: string | undefined;
+export const VERSION: string =
+  typeof __PACKAGE_VERSION__ !== 'undefined'
+    ? __PACKAGE_VERSION__
+    : '0.0.0-test';