diff --git a/qwen3-tts-backend/api/voice_designs.py b/qwen3-tts-backend/api/voice_designs.py
index a5e885a..059c440 100644
--- a/qwen3-tts-backend/api/voice_designs.py
+++ b/qwen3-tts-backend/api/voice_designs.py
@@ -81,6 +81,95 @@ async def list_voice_designs(
     total = crud.count_voice_designs(db, current_user.id, backend_type)
     return VoiceDesignListResponse(designs=[to_voice_design_response(d) for d in designs], total=total)
 
+@router.post("/prepare-and-create", response_model=VoiceDesignResponse, status_code=status.HTTP_201_CREATED)
+@limiter.limit("10/minute")
+async def prepare_and_create_voice_design(
+    request: Request,
+    data: VoiceDesignCreate,
+    current_user: User = Depends(get_current_user),
+    db: Session = Depends(get_db)
+):
+    from core.tts_service import TTSServiceFactory
+    from core.cache_manager import VoiceCacheManager
+    from utils.audio import process_ref_audio, extract_audio_features
+    from core.config import settings
+    from db.crud import can_user_use_local_model
+    from datetime import datetime
+
+    if not can_user_use_local_model(current_user):
+        raise HTTPException(status_code=403, detail="Local model access required")
+
+    try:
+        backend = await TTSServiceFactory.get_backend("local")
+        ref_text = data.preview_text or data.instruct[:100]
+
+        ref_audio_bytes, _ = await backend.generate_voice_design({
+            "text": ref_text,
+            "language": "Auto",
+            "instruct": data.instruct,
+            "max_new_tokens": 2048,
+            "temperature": 0.3,
+            "top_k": 10,
+            "top_p": 0.5,
+            "repetition_penalty": 1.05
+        })
+
+        timestamp = datetime.utcnow().strftime("%Y%m%d_%H%M%S")
+        ref_filename = f"voice_design_new_{timestamp}.wav"
+        ref_audio_path = Path(settings.OUTPUT_DIR) / ref_filename
+        with open(ref_audio_path, 'wb') as f:
+            f.write(ref_audio_bytes)
+
+        ref_audio_array, ref_sr = process_ref_audio(ref_audio_bytes)
+
+        from core.model_manager import ModelManager
+        model_manager = await ModelManager.get_instance()
+        await model_manager.load_model("base")
+        _, tts = await model_manager.get_current_model()
+        if tts is None:
+            raise RuntimeError("Failed to load base model")
+
+        x_vector = tts.create_voice_clone_prompt(
+            ref_audio=(ref_audio_array, ref_sr),
+            ref_text=ref_text,
+            x_vector_only_mode=True
+        )
+
+        cache_manager = await VoiceCacheManager.get_instance()
+        ref_audio_hash = cache_manager.get_audio_hash(ref_audio_bytes)
+        features = extract_audio_features(ref_audio_array, ref_sr)
+        metadata = {
+            'duration': features['duration'],
+            'sample_rate': features['sample_rate'],
+            'ref_text': ref_text,
+            'x_vector_only_mode': True,
+            'instruct': data.instruct
+        }
+        cache_id = await cache_manager.set_cache(
+            current_user.id, ref_audio_hash, x_vector, metadata, db
+        )
+
+        design = crud.create_voice_design(
+            db=db,
+            user_id=current_user.id,
+            name=data.name,
+            instruct=data.instruct,
+            backend_type="local",
+            meta_data=data.meta_data,
+            preview_text=data.preview_text,
+            voice_cache_id=cache_id,
+            ref_audio_path=str(ref_audio_path),
+            ref_text=ref_text,
+        )
+
+        logger.info(f"Voice design created with clone prompt: design_id={design.id}, cache_id={cache_id}")
+        return to_voice_design_response(design)
+
+    except Exception as e:
+        logger.error(f"Failed to prepare and create voice design: {e}", exc_info=True)
+        raise HTTPException(status_code=500, detail="Failed to prepare voice design")
+
+
 @router.post("/{design_id}/prepare-clone")
 @limiter.limit("10/minute")
 async def prepare_voice_clone_prompt(
diff --git a/qwen3-tts-backend/db/crud.py b/qwen3-tts-backend/db/crud.py
index 442aa2b..8768a0d 100644
--- a/qwen3-tts-backend/db/crud.py
+++ b/qwen3-tts-backend/db/crud.py
@@ -280,7 +280,10 @@ def create_voice_design(
     backend_type: str,
     aliyun_voice_id: Optional[str] = None,
     meta_data: Optional[Dict[str, Any]] = None,
-    preview_text: Optional[str] = None
+    preview_text: Optional[str] = None,
+    voice_cache_id: Optional[int] = None,
+    ref_audio_path: Optional[str] = None,
+    ref_text: Optional[str] = None,
 ) -> VoiceDesign:
     design = VoiceDesign(
         user_id=user_id,
@@ -290,6 +293,9 @@ def create_voice_design(
         aliyun_voice_id=aliyun_voice_id,
         meta_data=meta_data,
         preview_text=preview_text,
+        voice_cache_id=voice_cache_id,
+        ref_audio_path=ref_audio_path,
+        ref_text=ref_text,
         created_at=datetime.utcnow(),
         last_used=datetime.utcnow()
     )
diff --git a/qwen3-tts-frontend/src/components/tts/VoiceDesignForm.tsx b/qwen3-tts-frontend/src/components/tts/VoiceDesignForm.tsx
index 251cd4e..ab06424 100644
--- a/qwen3-tts-frontend/src/components/tts/VoiceDesignForm.tsx
+++ b/qwen3-tts-frontend/src/components/tts/VoiceDesignForm.tsx
@@ -145,28 +145,27 @@ const VoiceDesignForm = forwardRef<VoiceDesignFormHandle>((_props, ref) => {
       return
     }
 
+    const backend = preferences?.default_backend || 'local'
+    const text = watch('text')
+    const designData = {
+      name: saveDesignName,
+      instruct: instruct,
+      backend_type: backend,
+      preview_text: text || `${saveDesignName}的声音`
+    }
+
     try {
-      const backend = preferences?.default_backend || 'local'
-      const text = watch('text')
-      const design = await voiceDesignApi.create({
-        name: saveDesignName,
-        instruct: instruct,
-        backend_type: backend,
-        preview_text: text || `${saveDesignName}的声音`
-      })
-
-      toast.success(t('designSaved'))
-
       if (backend === 'local') {
         setIsPreparing(true)
         try {
-          await voiceDesignApi.prepareClone(design.id)
-          toast.success(t('clonePrepared'))
-        } catch (error) {
-          toast.error(t('clonePrepareFailed'))
+          await voiceDesignApi.prepareAndCreate(designData)
+          toast.success(t('designSaved'))
         } finally {
           setIsPreparing(false)
         }
+      } else {
+        await voiceDesignApi.create(designData)
+        toast.success(t('designSaved'))
       }
 
       setShowSaveDialog(false)
diff --git a/qwen3-tts-frontend/src/lib/api.ts b/qwen3-tts-frontend/src/lib/api.ts
index aea6a1f..e11a2d2 100644
--- a/qwen3-tts-frontend/src/lib/api.ts
+++ b/qwen3-tts-frontend/src/lib/api.ts
@@ -432,6 +432,14 @@ export const voiceDesignApi = {
     return response.data
   },
 
+  prepareAndCreate: async (data: VoiceDesignCreate): Promise<VoiceDesign> => {
+    const response = await apiClient.post<VoiceDesign>(
+      API_ENDPOINTS.VOICE_DESIGNS.PREPARE_AND_CREATE,
+      data
+    )
+    return response.data
+  },
+
   prepareClone: async (id: number): Promise<{ message: string; cache_id: number; ref_text: string }> => {
     const response = await apiClient.post<{ message: string; cache_id: number; ref_text: string }>(
       API_ENDPOINTS.VOICE_DESIGNS.PREPARE_CLONE(id)
diff --git a/qwen3-tts-frontend/src/lib/constants.ts b/qwen3-tts-frontend/src/lib/constants.ts
index 105a14f..afe43e9 100644
--- a/qwen3-tts-frontend/src/lib/constants.ts
+++ b/qwen3-tts-frontend/src/lib/constants.ts
@@ -30,6 +30,7 @@ export const API_ENDPOINTS = {
   VOICE_DESIGNS: {
     LIST: '/voice-designs',
     CREATE: '/voice-designs',
+    PREPARE_AND_CREATE: '/voice-designs/prepare-and-create',
     PREPARE_CLONE: (id: number) => `/voice-designs/${id}/prepare-clone`,
   },
 } as const