feat(asr): refactor OpenAI ASR integration, add Deepgram/OpenAI ASR adapters, and polish provider tests

aisuite-js/README.md

@@ -28,8 +28,7 @@ import { Client } from 'aisuite';
 
 const client = new Client({
   openai: { 
-    apiKey: process.env.OPENAI_API_KEY,
-    audio: true  // Enable Whisper ASR support
+    apiKey: process.env.OPENAI_API_KEY,    
   },
   anthropic: { apiKey: process.env.ANTHROPIC_API_KEY },
   deepgram: { apiKey: process.env.DEEPGRAM_API_KEY },
@@ -154,8 +153,7 @@ try {
 // Initialize client with audio support for OpenAI
 const client = new Client({
   openai: { 
-    apiKey: process.env.OPENAI_API_KEY,
-    audio: true  // Required for Whisper ASR
+    apiKey: process.env.OPENAI_API_KEY,    
   },
   deepgram: { apiKey: process.env.DEEPGRAM_API_KEY }
 });
@@ -170,7 +168,7 @@ const deepgramResponse = await client.audio.transcriptions.create({
   speaker_labels: true,
 });
 
-// Using OpenAI Whisper (requires audio: true in config)
+// Using OpenAI Whisper
 const openaiResponse = await client.audio.transcriptions.create({
   model: 'openai:whisper-1',
   file: audioBuffer,
@@ -214,8 +212,7 @@ const client = new Client({
   openai?: {
     apiKey: string;
     baseURL?: string;
-    organization?: string;
-    audio?: boolean;      // Enable Whisper ASR support
+    organization?: string;    
   },
   anthropic?: {
     apiKey: string;
@@ -280,7 +277,6 @@ client.isASRProviderConfigured('deepgram'); // true
 - Streaming tool calls require manual accumulation of arguments
 - ASR support is limited to OpenAI Whisper (requires explicit audio configuration) and Deepgram
 - Some provider-specific ASR features might require using provider-specific parameters
-- OpenAI Whisper support requires additional `audio: true` configuration
 
 ## Development
 

aisuite-js/examples/openai-asr.ts

@@ -6,8 +6,7 @@ async function main() {
   // Initialize the client with OpenAI configuration
   const client = new Client({
     openai: {
-      apiKey: process.env.OPENAI_API_KEY || "your-openai-api-key",
-      audio: true,
+      apiKey: process.env.OPENAI_API_KEY!, 
     },
   });
 

aisuite-js/src/asr-providers/deepgram/provider.ts

@@ -1,5 +1,5 @@
 import { createClient, DeepgramClient } from "@deepgram/sdk";
-import { BaseASRProvider } from "../../core/base-asr-provider";
+import { ASRProvider } from "../../core/base-asr-provider";
 import {
   TranscriptionRequest,
   TranscriptionResult,
@@ -10,13 +10,11 @@ import { adaptResponse } from "./adapters";
 import { AISuiteError } from "../../core/errors";
 import * as fs from "fs";
 
-export class DeepgramASRProvider extends BaseASRProvider {
+export class DeepgramASRProvider implements ASRProvider {
   public readonly name = "deepgram";
   private client: DeepgramClient;
 
   constructor(config: DeepgramConfig) {
-    super();
-
     // Use the new createClient API instead of the deprecated Deepgram constructor
     this.client = createClient({
       key: config.apiKey,

aisuite-js/src/asr-providers/index.ts

@@ -1,6 +1,2 @@
 export { DeepgramASRProvider } from "./deepgram";
-export { OpenAIASRProvider } from "./openai";
-export type {
-  OpenAIConfig,
-  DeepgramConfig,
-} from "../types";
+export type { DeepgramConfig } from "../types";

aisuite-js/src/client.ts

@@ -7,32 +7,29 @@ import {
   TranscriptionRequest,
   TranscriptionResult,
 } from "./types";
-import { BaseProvider } from "./core/base-provider";
-import { BaseASRProvider } from "./core/base-asr-provider";
+import { Provider } from "./core/base-provider";
+import { ASRProvider } from "./core/base-asr-provider";
 import { parseModel } from "./core/model-parser";
 import { ProviderNotConfiguredError } from "./core/errors";
 import { OpenAIProvider } from "./providers/openai";
 import { AnthropicProvider } from "./providers/anthropic";
 import { MistralProvider } from "./providers/mistral";
 import { GroqProvider } from "./providers/groq";
 import { DeepgramASRProvider } from "./asr-providers/deepgram";
-import { OpenAIASRProvider } from "./asr-providers/openai";
 
 export class Client {
-  private chatProviders: Map<string, BaseProvider> = new Map();
-  private asrProviders: Map<string, BaseASRProvider> = new Map();
+  private chatProviders: Map<string, Provider> = new Map();
+  private asrProviders: Map<string, ASRProvider> = new Map();
 
   constructor(config: ProviderConfigs) {
     this.initializeProviders(config);
   }
 
   private initializeProviders(config: ProviderConfigs): void {
     if (config.openai) {
-      if (config.openai.audio) {
-        this.asrProviders.set("openai", new OpenAIASRProvider(config.openai));
-      } else {
-        this.chatProviders.set("openai", new OpenAIProvider(config.openai));
-      }
+      const openaiProvider = new OpenAIProvider(config.openai);
+      this.chatProviders.set("openai", openaiProvider);
+      this.asrProviders.set("openai", openaiProvider);
     }
 
     if (config.anthropic) {

aisuite-js/src/core/base-asr-provider.ts

@@ -4,19 +4,19 @@ import {
   RequestOptions 
 } from '../types';
 
-export abstract class BaseASRProvider {
-  abstract readonly name: string;
+export interface ASRProvider {
+  readonly name: string;
 
-  abstract transcribe(
+  transcribe(
     request: TranscriptionRequest,
     options?: RequestOptions
   ): Promise<TranscriptionResult>;
 
-  abstract validateParams(    
+  validateParams(    
     params: { [key: string]: any }
   ): void;
 
-  abstract translateParams(    
+  translateParams(    
     params: { [key: string]: any }
   ): { [key: string]: any };
 }

aisuite-js/src/core/base-provider.ts

@@ -5,15 +5,15 @@ import {
   RequestOptions 
 } from '../types';
 
-export abstract class BaseProvider {
-  abstract readonly name: string;
+export interface Provider {
+  readonly name: string;
 
-  abstract chatCompletion(
+  chatCompletion(
     request: ChatCompletionRequest,
     options?: RequestOptions
   ): Promise<ChatCompletionResponse>;
 
-  abstract streamChatCompletion(
+  streamChatCompletion(
     request: ChatCompletionRequest,
     options?: RequestOptions
   ): AsyncIterable<ChatCompletionChunk>;

aisuite-js/src/index.ts

@@ -11,7 +11,4 @@ export {
   MistralProvider,
 } from "./providers";
 
-export {
-  DeepgramASRProvider,
-  OpenAIASRProvider,
-} from "./asr-providers";
+export { DeepgramASRProvider } from "./asr-providers";

aisuite-js/src/providers/anthropic/provider.ts

@@ -1,5 +1,5 @@
 import Anthropic from '@anthropic-ai/sdk';
-import { BaseProvider } from '../../core/base-provider';
+import { Provider } from '../../core/base-provider';
 import { 
   ChatCompletionRequest, 
   ChatCompletionResponse, 
@@ -11,12 +11,11 @@ import { adaptRequest, adaptResponse, adaptStreamEvent } from './adapters';
 import { AISuiteError } from '../../core/errors';
 import { generateId } from '../../utils/streaming';
 
-export class AnthropicProvider extends BaseProvider {
+export class AnthropicProvider implements Provider {
   public readonly name = 'anthropic';
   private client: Anthropic;
 
   constructor(config: AnthropicConfig) {
-    super();
     this.client = new Anthropic({
       apiKey: config.apiKey,
       baseURL: config.baseURL,

aisuite-js/src/providers/groq/provider.ts

@@ -1,5 +1,5 @@
 import Groq from "groq-sdk";
-import { BaseProvider } from "../../core/base-provider";
+import { Provider } from "../../core/base-provider";
 import {
   ChatCompletionRequest,
   ChatCompletionResponse,
@@ -11,13 +11,11 @@ import { adaptRequest, adaptResponse, adaptStreamResponse } from "./adapters";
 import { AISuiteError } from "../../core/errors";
 import { generateId } from "../../utils/streaming";
 
-export class GroqProvider extends BaseProvider {
+export class GroqProvider implements Provider {
   public readonly name = "groq";
   private client: Groq;
 
   constructor(config: GroqConfig) {
-    super();
-
     this.client = new Groq({
       apiKey: config.apiKey,
       dangerouslyAllowBrowser: config.dangerouslyAllowBrowser || false, // Allow browser usage for chat app