added elevenlabs stt support (#222)

* added elevenlabs stt support

* review comment fix

Author: rootflo-hardik

wavefront/client/src/config/voice-providers.ts

@@ -220,7 +220,7 @@ export const VOICE_PROVIDERS_CONFIG: VoiceProvidersConfig = {
     },
   },
   stt: {
-    providers: ['deepgram', 'sarvam'] as const,
+    providers: ['deepgram', 'sarvam', 'elevenlabs'] as const,
     configs: {
       deepgram: {
         name: 'Deepgram',
@@ -325,6 +325,33 @@ export const VOICE_PROVIDERS_CONFIG: VoiceProvidersConfig = {
           },
         },
       },
+      elevenlabs: {
+        name: 'ElevenLabs',
+        badge: {
+          bg: 'bg-purple-100',
+          text: 'text-purple-800',
+        },
+        parameters: {
+          model: {
+            type: 'string',
+            default: 'scribe_v2_realtime',
+            description: 'ElevenLabs STT model',
+            options: ['scribe_v2_realtime'],
+          },
+          language: {
+            type: 'string',
+            default: '',
+            description: 'Language code (ISO-639-1, e.g., en, hi)',
+            placeholder: 'en',
+          },
+          sample_rate: {
+            type: 'number',
+            default: 8000,
+            description: 'Audio sample rate in Hz',
+            placeholder: '8000',
+          },
+        },
+      },
     },
   },
 };

wavefront/client/src/pages/apps/[appId]/voice-agents/stt-configs/CreateSttConfigDialog.tsx

@@ -22,6 +22,7 @@ import { Input } from '@app/components/ui/input';
 import { Select, SelectContent, SelectItem, SelectTrigger, SelectValue } from '@app/components/ui/select';
 import { Textarea } from '@app/components/ui/textarea';
 import { VOICE_PROVIDERS_CONFIG, getProviderConfig } from '@app/config/voice-providers';
+import { SttProvider } from '@app/types/stt-config';
 import { extractErrorMessage } from '@app/lib/utils';
 import { useNotifyStore } from '@app/store';
 import { zodResolver } from '@hookform/resolvers/zod';
@@ -32,7 +33,7 @@ import { z } from 'zod';
 const createSttConfigSchema = z.object({
   display_name: z.string().min(1, 'Display name is required').max(100, 'Display name must be 100 characters or less'),
   description: z.string().max(500, 'Description must be 500 characters or less').optional(),
-  provider: z.enum(['deepgram', 'sarvam'] as [string, ...string[]]),
+  provider: z.enum(['deepgram', 'sarvam', 'elevenlabs'] as [string, ...string[]]),
   api_key: z.string().min(1, 'API key is required'),
 });
 
@@ -76,7 +77,7 @@ const CreateSttConfigDialog: React.FC<CreateSttConfigDialogProps> = ({ isOpen, o
       await floConsoleService.sttConfigService.createSttConfig({
         display_name: data.display_name.trim(),
         description: data.description?.trim() || null,
-        provider: data.provider as 'deepgram',
+        provider: data.provider as SttProvider,
         api_key: data.api_key.trim(),
       });
       notifySuccess('STT configuration created successfully');

wavefront/client/src/pages/apps/[appId]/voice-agents/stt-configs/EditSttConfigDialog.tsx

@@ -33,7 +33,7 @@ import { z } from 'zod';
 const updateSttConfigSchema = z.object({
   display_name: z.string().min(1, 'Display name is required').max(100, 'Display name must be 100 characters or less'),
   description: z.string().max(500, 'Description must be 500 characters or less').optional(),
-  provider: z.enum(['deepgram', 'sarvam'] as [string, ...string[]]),
+  provider: z.enum(['deepgram', 'sarvam', 'elevenlabs'] as [string, ...string[]]),
   api_key: z.string().optional(),
 });
 

wavefront/client/src/types/stt-config.ts

@@ -1,6 +1,6 @@
 import { IApiResponse } from '@app/lib/axios';
 
-export type SttProvider = 'deepgram' | 'sarvam';
+export type SttProvider = 'deepgram' | 'sarvam' | 'elevenlabs';
 
 export interface SttConfig {
   id: string;
@@ -60,3 +60,10 @@ export interface SarvamSttParameters {
   vad_signals?: boolean;
   high_vad_sensitivity?: boolean;
 }
+
+// ElevenLabs STT specific parameters
+export interface ElevenLabsSttParameters {
+  model?: string; // default: 'scribe_v2_realtime'
+  language?: string; // ISO-639-1 code e.g. 'en', 'hi'
+  sample_rate?: number; // default: 8000
+}

wavefront/server/apps/call_processing/call_processing/services/stt_service.py

@@ -1,7 +1,7 @@
 """
 STT (Speech-to-Text) service factory
 
-Supports multiple providers: Deepgram, AssemblyAI, Whisper, Google, Azure
+Supports multiple providers: Deepgram, Sarvam, ElevenLabs
 """
 
 from typing import Dict, Any
@@ -10,6 +10,7 @@
 # Pipecat STT services
 from pipecat.services.deepgram.stt import DeepgramSTTService
 from pipecat.services.sarvam.stt import SarvamSTTService
+from pipecat.services.elevenlabs.stt import ElevenLabsRealtimeSTTService
 
 # Pipecat language enum
 from pipecat.transcriptions.language import Language
@@ -57,6 +58,8 @@ def create_stt_service(stt_config: Dict[str, Any]):
             return STTServiceFactory._create_deepgram_stt(api_key, parameters)
         elif provider == 'sarvam':
             return STTServiceFactory._create_sarvam_stt(api_key, parameters)
+        elif provider == 'elevenlabs':
+            return STTServiceFactory._create_elevenlabs_stt(api_key, parameters)
         elif provider == 'assemblyai':
             return STTServiceFactory._create_assemblyai_stt(api_key, parameters)
         elif provider == 'whisper':
@@ -162,6 +165,55 @@ def _create_sarvam_stt(api_key: str, parameters: Dict[str, Any]):
             params=input_params,
         )
 
+    # Mapping of short language codes to ElevenLabs ISO-639-3 language codes
+    ELEVENLABS_LANGUAGE_MAP = {
+        'en': 'eng',
+        'hi': 'hin',
+        'ta': 'tam',
+        'te': 'tel',
+        'kn': 'kan',
+        'ml': 'mal',
+        'gu': 'guj',
+        'bn': 'ben',
+        'mr': 'mar',
+        'pa': 'pan',
+        'or': 'ori',
+    }
+
+    @staticmethod
+    def _create_elevenlabs_stt(api_key: str, parameters: Dict[str, Any]):
+        """Create ElevenLabs Realtime STT service (WebSocket streaming, scribe_v2_realtime)"""
+        params_dict = {}
+
+        # Map language code to ElevenLabs ISO-639-3 code
+        if 'language' in parameters and parameters['language']:
+            lang_code = parameters['language']
+            elevenlabs_lang = STTServiceFactory.ELEVENLABS_LANGUAGE_MAP.get(lang_code)
+            if elevenlabs_lang:
+                params_dict['language_code'] = elevenlabs_lang
+            else:
+                logger.warning(
+                    f"Unknown ElevenLabs language '{lang_code}', skipping (auto-detect will be used)"
+                )
+
+        model = parameters.get('model', 'scribe_v2_realtime')
+        sample_rate = parameters.get('sample_rate', 8000)
+
+        input_params = (
+            ElevenLabsRealtimeSTTService.InputParams(**params_dict)
+            if params_dict
+            else None
+        )
+
+        logger.info(f'ElevenLabs STT config: model={model}, sample_rate={sample_rate}')
+
+        return ElevenLabsRealtimeSTTService(
+            api_key=api_key,
+            model=model,
+            sample_rate=sample_rate,
+            params=input_params,
+        )
+
     @staticmethod
     def _create_assemblyai_stt(api_key: str, parameters: Dict[str, Any]):
         """Create AssemblyAI STT service"""

wavefront/server/modules/voice_agents_module/voice_agents_module/models/stt_schemas.py

@@ -15,6 +15,7 @@ class SttProvider(str, Enum):
     GOOGLE = 'google'
     AZURE = 'azure'
     SARVAM = 'sarvam'
+    ELEVENLABS = 'elevenlabs'
 
 
 class CreateSttConfigPayload(BaseModel):