diff --git a/mnemo_cards_admin/src/api/voiceService.ts b/mnemo_cards_admin/src/api/voiceService.ts
index 64cbf17..6e68373 100644
--- a/mnemo_cards_admin/src/api/voiceService.ts
+++ b/mnemo_cards_admin/src/api/voiceService.ts
@@ -11,17 +11,20 @@ export const voiceServiceApi = {
    * Backend calls voice-service, which handles caching, and saves to MinIO
    * @param text Text to convert to speech
    * @param voice Voice identifier (optional, defaults to 'alloy')
+   * @param language Language code (optional)
    * @returns Object ID and presigned URL
    */
   async generateTTS(
     text: string,
-    voice?: string
+    voice?: string,
+    language?: string
   ): Promise<GenerateTTSResponse> {
     const response = await adminApiClient.post<GenerateTTSResponse>(
       '/api/v2/media/generate-tts',
       {
         text,
         voice: voice || 'alloy',
+        ...(language && { language }),
       }
     )
 
diff --git a/mnemo_cards_admin/src/components/CardVoicesManager.tsx b/mnemo_cards_admin/src/components/CardVoicesManager.tsx
index 98a7a09..08c4997 100644
--- a/mnemo_cards_admin/src/components/CardVoicesManager.tsx
+++ b/mnemo_cards_admin/src/components/CardVoicesManager.tsx
@@ -24,7 +24,7 @@ export function CardVoicesManager({
   const queryClient = useQueryClient()
   const [showAddForm, setShowAddForm] = useState(false)
   const [newAudio, setNewAudio] = useState<string | undefined>(undefined)
-  const [newLanguage, setNewLanguage] = useState('en')
+  const [newLanguage, setNewLanguage] = useState('es')
   const [isGeneratingTTS, setIsGeneratingTTS] = useState(false)
 
   // Load voices for the card
@@ -65,7 +65,7 @@ export function CardVoicesManager({
       // Auto-close form and reset after successful addition
       setShowAddForm(false)
       setNewAudio(undefined)
-      setNewLanguage('en')
+      setNewLanguage('es')
     },
     onError: (error: unknown) => {
       console.error('❌ CardVoicesManager: addVoiceMutation.onError', error)
@@ -111,14 +111,14 @@ export function CardVoicesManager({
     try {
       const text = cardOriginal.trim()
       // Use default voice 'alloy' for TTS generation
-      // Language (newLanguage) is used only when saving the voice to the card
+      // Language (newLanguage) is used for both TTS generation and saving the voice to the card
       const voice = 'alloy'
       
-      console.log('🔄 CardVoicesManager: Requesting TTS generation', { text, voice })
+      console.log('🔄 CardVoicesManager: Requesting TTS generation', { text, voice, language: newLanguage })
       
       // Generate TTS via backend API
       // Backend calls voice-service (which handles caching) and saves to MinIO
-      const ttsResponse = await voiceServiceApi.generateTTS(text, voice)
+      const ttsResponse = await voiceServiceApi.generateTTS(text, voice, newLanguage)
       console.log('✅ CardVoicesManager: TTS generated and saved to MinIO', { 
         objectId: ttsResponse.objectId 
       })
@@ -126,7 +126,7 @@ export function CardVoicesManager({
       // Add voice to card using the objectId from backend
       await addVoiceMutation.mutateAsync({
         voiceUrl: ttsResponse.objectId,
-        language: newLanguage || 'en',
+        language: newLanguage || 'es',
       })
     } catch (error) {
       console.error('❌ CardVoicesManager: TTS generation failed', error)
@@ -224,7 +224,7 @@ export function CardVoicesManager({
                 onClick={() => {
                   setShowAddForm(false)
                   setNewAudio(undefined)
-                  setNewLanguage('en')
+                  setNewLanguage('es')
                 }}
                 disabled={disabled || addVoiceMutation.isPending}
               >
diff --git a/mnemo_cards_backend/lib/api/v2/media_api_v2.dart b/mnemo_cards_backend/lib/api/v2/media_api_v2.dart
index a1fe80d..8635d5b 100644
--- a/mnemo_cards_backend/lib/api/v2/media_api_v2.dart
+++ b/mnemo_cards_backend/lib/api/v2/media_api_v2.dart
@@ -508,13 +508,14 @@ class MediaApiV2 {
       final json = jsonDecode(body) as Map<String, dynamic>;
       final text = json['text'] as String?;
       final voice = json['voice'] as String?;
+      final language = json['language'] as String?;
 
       if (text == null || text.trim().isEmpty) {
         return _badRequest('Text is required');
       }
 
       print(
-        '🔄 MediaApiV2: Requesting TTS from voice-service: text="${text.substring(0, text.length > 50 ? 50 : text.length)}", voice="${voice ?? 'default'}"',
+        '🔄 MediaApiV2: Requesting TTS from voice-service: text="${text.substring(0, text.length > 50 ? 50 : text.length)}", voice="${voice ?? 'default'}", language="${language ?? 'not provided'}"',
       );
 
       // Call voice-service
@@ -524,6 +525,7 @@ class MediaApiV2 {
           'text': text,
           if (voice != null) 'voice': voice,
           'format': 'mp3',
+          if (language != null) 'language': language,
         },
         options: dio.Options(
           responseType: dio.ResponseType.bytes,