diff --git a/services/search.py b/services/search.py
index 498cf8e4..77330c6a 100644
--- a/services/search.py
+++ b/services/search.py
@@ -6,7 +6,7 @@ from typing import Any, Dict, List
 import muvera
 import numpy as np
 
-from settings import SEARCH_MAX_BATCH_SIZE, SEARCH_PREFETCH_SIZE
+from settings import MUVERA_INDEX_NAME, SEARCH_MAX_BATCH_SIZE, SEARCH_PREFETCH_SIZE
 from utils.logger import root_logger as logger
 
 # Global collection for background tasks
@@ -34,13 +34,13 @@ class MuveraWrapper:
         }
 
     async def search(self, query: str, limit: int) -> List[Dict[str, Any]]:
-        """Simple search implementation using FDE encoding"""
+        """Simple search implementation using FDE encoding with deterministic results"""
         if not query.strip():
             return []
 
-        # For demo purposes, create a simple query embedding
-        # In a real implementation, you'd use a proper text embedding model
-        rng = np.random.default_rng()
+        # Create deterministic query embedding based on query hash
+        query_hash = hash(query.strip().lower())
+        rng = np.random.default_rng(seed=query_hash & 0x7FFFFFFF)  # Use positive seed
         query_embedding = rng.standard_normal((32, self.vector_dimension)).astype(np.float32)
 
         # Encode query using FDE
@@ -62,19 +62,20 @@ class MuveraWrapper:
                     }
                 )
 
-        # Sort by score and limit results
-        results.sort(key=lambda x: x["score"], reverse=True)
+        # Sort by score and limit results - добавляем сортировку по ID для стабильности
+        results.sort(key=lambda x: (x["score"], x["id"]), reverse=True)
         return results[:limit]
 
     async def index(self, documents: List[Dict[str, Any]]) -> None:
-        """Index documents using FDE encoding"""
+        """Index documents using FDE encoding with deterministic embeddings"""
         for doc in documents:
             doc_id = doc["id"]
             self.documents[doc_id] = doc
 
-            # Create a simple document embedding (in real implementation, use proper text embedding)
-            # For now, create random embeddings for demo
-            rng = np.random.default_rng()
+            # Create deterministic document embedding based on document content hash
+            doc_content = f"{doc.get('title', '')} {doc.get('body', '')}"
+            content_hash = hash(doc_content.strip().lower() + str(doc_id))
+            rng = np.random.default_rng(seed=content_hash & 0x7FFFFFFF)  # Use positive seed
             doc_embedding = rng.standard_normal((32, self.vector_dimension)).astype(np.float32)
 
             # Encode document using FDE (average aggregation for documents)
@@ -104,16 +105,16 @@ class SearchService:
         self.muvera_client: Any = None
         self.client: Any = None
 
-        # Initialize Muvera
+        # Initialize local Muvera
         try:
-            # Initialize Muvera wrapper with your configuration
             self.muvera_client = MuveraWrapper(
                 vector_dimension=768,  # Standard embedding dimension
                 cache_enabled=True,
                 batch_size=SEARCH_MAX_BATCH_SIZE,
             )
             self.available = True
-            logger.info("Muvera wrapper initialized successfully - enhanced search enabled")
+            logger.info(f"Local Muvera wrapper initialized - index: {MUVERA_INDEX_NAME}")
+
         except Exception as e:
             logger.error(f"Failed to initialize Muvera: {e}")
             self.available = False
@@ -126,7 +127,7 @@ class SearchService:
             # Get Muvera service info
             if self.muvera_client:
                 muvera_info = await self.muvera_client.info()
-                return {"status": "enabled", "provider": "muvera", "muvera_info": muvera_info}
+                return {"status": "enabled", "provider": "muvera", "mode": "local", "muvera_info": muvera_info}
             return {"status": "error", "message": "Muvera client not available"}
         except Exception:
             logger.exception("Failed to get search info")
@@ -403,6 +404,7 @@ class SearchService:
         if hasattr(self, "muvera_client") and self.muvera_client:
             try:
                 await self.muvera_client.close()
+                logger.info("Local Muvera client closed")
             except Exception as e:
                 logger.warning(f"Error closing Muvera client: {e}")
         logger.info("Search service closed")
diff --git a/settings.py b/settings.py
index e91ecc6b..57116575 100644
--- a/settings.py
+++ b/settings.py
@@ -96,3 +96,4 @@ SEARCH_MAX_BATCH_SIZE = int(os.environ.get("SEARCH_MAX_BATCH_SIZE", "25"))
 SEARCH_CACHE_ENABLED = bool(os.environ.get("SEARCH_CACHE_ENABLED", "true").lower() in ["true", "1", "yes"])
 SEARCH_CACHE_TTL_SECONDS = int(os.environ.get("SEARCH_CACHE_TTL_SECONDS", "300"))
 SEARCH_PREFETCH_SIZE = int(os.environ.get("SEARCH_PREFETCH_SIZE", "200"))
+MUVERA_INDEX_NAME = "discours"