@ai-sdk/revai 3.0.0-beta.4 → 3.0.0-beta.51

package/src/revai-transcription-model.ts

@@ -1,7 +1,7 @@
 import {
   AISDKError,
-  TranscriptionModelV4,
-  SharedV4Warning,
+  type TranscriptionModelV4,
+  type SharedV4Warning,
 } from '@ai-sdk/provider';
 import {
   combineHeaders,
@@ -12,207 +12,16 @@ import {
   getFromApi,
   parseProviderOptions,
   postFormDataToApi,
+  serializeModelOptions,
+  WORKFLOW_SERIALIZE,
+  WORKFLOW_DESERIALIZE,
 } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
-import { RevaiConfig } from './revai-config';
+import type { RevaiConfig } from './revai-config';
 import { revaiFailedResponseHandler } from './revai-error';
-import { RevaiTranscriptionModelId } from './revai-transcription-options';
-import { RevaiTranscriptionAPITypes } from './revai-api-types';
-
-// https://docs.rev.ai/api/asynchronous/reference/#operation/SubmitTranscriptionJob
-const revaiTranscriptionModelOptionsSchema = z.object({
-  /**
-   * Optional metadata string to associate with the transcription job.
-   */
-  metadata: z.string().nullish(),
-  /**
-   * Configuration for webhook notifications when job is complete.
-   */
-  notification_config: z
-    .object({
-      /**
-       * URL to send the notification to.
-       */
-      url: z.string(),
-      /**
-       * Optional authorization headers for the notification request.
-       */
-      auth_headers: z
-        .object({
-          Authorization: z.string(),
-        })
-        .nullish(),
-    })
-    .nullish(),
-  /**
-   * Number of seconds after which the job will be automatically deleted.
-   */
-  delete_after_seconds: z.number().nullish(),
-  /**
-   * Whether to include filler words and false starts in the transcription.
-   */
-  verbatim: z.boolean().optional(),
-  /**
-   * Whether to prioritize the job for faster processing.
-   */
-  rush: z.boolean().nullish().default(false),
-  /**
-   * Whether to run the job in test mode.
-   */
-  test_mode: z.boolean().nullish().default(false),
-  /**
-   * Specific segments of the audio to transcribe.
-   */
-  segments_to_transcribe: z
-    .array(
-      z.object({
-        /**
-         * Start time of the segment in seconds.
-         */
-        start: z.number(),
-        /**
-         * End time of the segment in seconds.
-         */
-        end: z.number(),
-      }),
-    )
-    .nullish(),
-  /**
-   * Names to assign to speakers in the transcription.
-   */
-  speaker_names: z
-    .array(
-      z.object({
-        /**
-         * Display name for the speaker.
-         */
-        display_name: z.string(),
-      }),
-    )
-    .nullish(),
-  /**
-   * Whether to skip speaker diarization.
-   */
-  skip_diarization: z.boolean().nullish().default(false),
-  /**
-   * Whether to skip post-processing steps.
-   */
-  skip_postprocessing: z.boolean().nullish().default(false),
-  /**
-   * Whether to skip adding punctuation to the transcription.
-   */
-  skip_punctuation: z.boolean().nullish().default(false),
-  /**
-   * Whether to remove disfluencies (um, uh, etc.) from the transcription.
-   */
-  remove_disfluencies: z.boolean().nullish().default(false),
-  /**
-   * Whether to remove atmospheric sounds from the transcription.
-   */
-  remove_atmospherics: z.boolean().nullish().default(false),
-  /**
-   * Whether to filter profanity from the transcription.
-   */
-  filter_profanity: z.boolean().nullish().default(false),
-  /**
-   * Number of speaker channels in the audio.
-   */
-  speaker_channels_count: z.number().nullish(),
-  /**
-   * Expected number of speakers in the audio.
-   */
-  speakers_count: z.number().nullish(),
-  /**
-   * Type of diarization to use.
-   */
-  diarization_type: z
-    .enum(['standard', 'premium'])
-    .nullish()
-    .default('standard'),
-  /**
-   * ID of a custom vocabulary to use for the transcription.
-   */
-  custom_vocabulary_id: z.string().nullish(),
-  /**
-   * Custom vocabularies to use for the transcription.
-   */
-  custom_vocabularies: z.array(z.object({})).optional(),
-  /**
-   * Whether to strictly enforce custom vocabulary.
-   */
-  strict_custom_vocabulary: z.boolean().optional(),
-  /**
-   * Configuration for generating a summary of the transcription.
-   */
-  summarization_config: z
-    .object({
-      /**
-       * Model to use for summarization.
-       */
-      model: z.enum(['standard', 'premium']).nullish().default('standard'),
-      /**
-       * Format of the summary.
-       */
-      type: z.enum(['paragraph', 'bullets']).nullish().default('paragraph'),
-      /**
-       * Custom prompt for the summarization.
-       */
-      prompt: z.string().nullish(),
-    })
-    .nullish(),
-  /**
-   * Configuration for translating the transcription.
-   */
-  translation_config: z
-    .object({
-      /**
-       * Target languages for translation.
-       */
-      target_languages: z.array(
-        z.object({
-          /**
-           * Language code for translation target.
-           */
-          language: z.enum([
-            'en',
-            'en-us',
-            'en-gb',
-            'ar',
-            'pt',
-            'pt-br',
-            'pt-pt',
-            'fr',
-            'fr-ca',
-            'es',
-            'es-es',
-            'es-la',
-            'it',
-            'ja',
-            'ko',
-            'de',
-            'ru',
-          ]),
-        }),
-      ),
-      /**
-       * Model to use for translation.
-       */
-      model: z.enum(['standard', 'premium']).nullish().default('standard'),
-    })
-    .nullish(),
-  /**
-   * Language of the audio content.
-   */
-  language: z.string().nullish().default('en'),
-  /**
-   * Whether to perform forced alignment.
-   */
-  forced_alignment: z.boolean().nullish().default(false),
-});
-
-export type RevaiTranscriptionModelOptions = z.infer<
-  typeof revaiTranscriptionModelOptionsSchema
->;
+import { revaiTranscriptionModelOptionsSchema } from './revai-transcription-model-options';
+import type { RevaiTranscriptionModelId } from './revai-transcription-options';
+import type { RevaiTranscriptionAPITypes } from './revai-api-types';
 
 interface RevaiTranscriptionModelConfig extends RevaiConfig {
   _internal?: {
@@ -227,6 +36,20 @@ export class RevaiTranscriptionModel implements TranscriptionModelV4 {
     return this.config.provider;
   }
 
+  static [WORKFLOW_SERIALIZE](model: RevaiTranscriptionModel) {
+    return serializeModelOptions({
+      modelId: model.modelId,
+      config: model.config,
+    });
+  }
+
+  static [WORKFLOW_DESERIALIZE](options: {
+    modelId: RevaiTranscriptionModelId;
+    config: RevaiTranscriptionModelConfig;
+  }) {
+    return new RevaiTranscriptionModel(options.modelId, options.config);
+  }
+
   constructor(
     readonly modelId: RevaiTranscriptionModelId,
     private readonly config: RevaiTranscriptionModelConfig,
@@ -324,7 +147,7 @@ export class RevaiTranscriptionModel implements TranscriptionModelV4 {
         path: '/speechtotext/v1/jobs',
         modelId: this.modelId,
       }),
-      headers: combineHeaders(this.config.headers(), options.headers),
+      headers: combineHeaders(this.config.headers?.(), options.headers),
       formData,
       failedResponseHandler: revaiFailedResponseHandler,
       successfulResponseHandler: createJsonResponseHandler(
@@ -364,7 +187,7 @@ export class RevaiTranscriptionModel implements TranscriptionModelV4 {
           path: `/speechtotext/v1/jobs/${jobId}`,
           modelId: this.modelId,
         }),
-        headers: combineHeaders(this.config.headers(), options.headers),
+        headers: combineHeaders(this.config.headers?.(), options.headers),
         failedResponseHandler: revaiFailedResponseHandler,
         successfulResponseHandler: createJsonResponseHandler(
           revaiTranscriptionJobResponseSchema,
@@ -398,7 +221,7 @@ export class RevaiTranscriptionModel implements TranscriptionModelV4 {
         path: `/speechtotext/v1/jobs/${jobId}/transcript`,
         modelId: this.modelId,
       }),
-      headers: combineHeaders(this.config.headers(), options.headers),
+      headers: combineHeaders(this.config.headers?.(), options.headers),
       failedResponseHandler: revaiFailedResponseHandler,
       successfulResponseHandler: createJsonResponseHandler(
         revaiTranscriptionResponseSchema,

package/dist/index.d.mts

@@ -1,148 +0,0 @@
-import { TranscriptionModelV4, ProviderV4 } from '@ai-sdk/provider';
-import { FetchFunction } from '@ai-sdk/provider-utils';
-import { z } from 'zod/v4';
-
-type RevaiConfig = {
-    provider: string;
-    url: (options: {
-        modelId: string;
-        path: string;
-    }) => string;
-    headers: () => Record<string, string | undefined>;
-    fetch?: FetchFunction;
-    generateId?: () => string;
-};
-
-type RevaiTranscriptionModelId = 'machine' | 'low_cost' | 'fusion';
-
-declare const revaiTranscriptionModelOptionsSchema: z.ZodObject<{
-    metadata: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    notification_config: z.ZodOptional<z.ZodNullable<z.ZodObject<{
-        url: z.ZodString;
-        auth_headers: z.ZodOptional<z.ZodNullable<z.ZodObject<{
-            Authorization: z.ZodString;
-        }, z.core.$strip>>>;
-    }, z.core.$strip>>>;
-    delete_after_seconds: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    verbatim: z.ZodOptional<z.ZodBoolean>;
-    rush: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    test_mode: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    segments_to_transcribe: z.ZodOptional<z.ZodNullable<z.ZodArray<z.ZodObject<{
-        start: z.ZodNumber;
-        end: z.ZodNumber;
-    }, z.core.$strip>>>>;
-    speaker_names: z.ZodOptional<z.ZodNullable<z.ZodArray<z.ZodObject<{
-        display_name: z.ZodString;
-    }, z.core.$strip>>>>;
-    skip_diarization: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    skip_postprocessing: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    skip_punctuation: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    remove_disfluencies: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    remove_atmospherics: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    filter_profanity: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-    speaker_channels_count: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    speakers_count: z.ZodOptional<z.ZodNullable<z.ZodNumber>>;
-    diarization_type: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-        standard: "standard";
-        premium: "premium";
-    }>>>>;
-    custom_vocabulary_id: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    custom_vocabularies: z.ZodOptional<z.ZodArray<z.ZodObject<{}, z.core.$strip>>>;
-    strict_custom_vocabulary: z.ZodOptional<z.ZodBoolean>;
-    summarization_config: z.ZodOptional<z.ZodNullable<z.ZodObject<{
-        model: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-            standard: "standard";
-            premium: "premium";
-        }>>>>;
-        type: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-            paragraph: "paragraph";
-            bullets: "bullets";
-        }>>>>;
-        prompt: z.ZodOptional<z.ZodNullable<z.ZodString>>;
-    }, z.core.$strip>>>;
-    translation_config: z.ZodOptional<z.ZodNullable<z.ZodObject<{
-        target_languages: z.ZodArray<z.ZodObject<{
-            language: z.ZodEnum<{
-                en: "en";
-                "en-us": "en-us";
-                "en-gb": "en-gb";
-                ar: "ar";
-                pt: "pt";
-                "pt-br": "pt-br";
-                "pt-pt": "pt-pt";
-                fr: "fr";
-                "fr-ca": "fr-ca";
-                es: "es";
-                "es-es": "es-es";
-                "es-la": "es-la";
-                it: "it";
-                ja: "ja";
-                ko: "ko";
-                de: "de";
-                ru: "ru";
-            }>;
-        }, z.core.$strip>>;
-        model: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodEnum<{
-            standard: "standard";
-            premium: "premium";
-        }>>>>;
-    }, z.core.$strip>>>;
-    language: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodString>>>;
-    forced_alignment: z.ZodDefault<z.ZodOptional<z.ZodNullable<z.ZodBoolean>>>;
-}, z.core.$strip>;
-type RevaiTranscriptionModelOptions = z.infer<typeof revaiTranscriptionModelOptionsSchema>;
-interface RevaiTranscriptionModelConfig extends RevaiConfig {
-    _internal?: {
-        currentDate?: () => Date;
-    };
-}
-declare class RevaiTranscriptionModel implements TranscriptionModelV4 {
-    readonly modelId: RevaiTranscriptionModelId;
-    private readonly config;
-    readonly specificationVersion = "v4";
-    get provider(): string;
-    constructor(modelId: RevaiTranscriptionModelId, config: RevaiTranscriptionModelConfig);
-    private getArgs;
-    doGenerate(options: Parameters<TranscriptionModelV4['doGenerate']>[0]): Promise<Awaited<ReturnType<TranscriptionModelV4['doGenerate']>>>;
-}
-
-interface RevaiProvider extends ProviderV4 {
-    (modelId: 'machine', settings?: {}): {
-        transcription: RevaiTranscriptionModel;
-    };
-    /**
-     * Creates a model for transcription.
-     */
-    transcription(modelId: RevaiTranscriptionModelId): TranscriptionModelV4;
-    /**
-     * @deprecated Use `embeddingModel` instead.
-     */
-    textEmbeddingModel(modelId: string): never;
-}
-interface RevaiProviderSettings {
-    /**
-     * 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 a Rev.ai provider instance.
- */
-declare function createRevai(options?: RevaiProviderSettings): RevaiProvider;
-/**
- * Default Rev.ai provider instance.
- */
-declare const revai: RevaiProvider;
-
-declare const VERSION: string;
-
-export { type RevaiProvider, type RevaiProviderSettings, type RevaiTranscriptionModelOptions, VERSION, createRevai, revai };