diff --git a/api/core/rag/datasource/vdb/weaviate/weaviate_vector.py b/api/core/rag/datasource/vdb/weaviate/weaviate_vector.py
index b48dd93f04..4cf7afa1f2 100644
--- a/api/core/rag/datasource/vdb/weaviate/weaviate_vector.py
+++ b/api/core/rag/datasource/vdb/weaviate/weaviate_vector.py
@@ -78,6 +78,7 @@ class WeaviateVector(BaseVector):
             attributes: List of metadata attributes to store
         """
         super().__init__(collection_name)
+        self._config = config
         self._client = self._init_client(config)
         self._attributes = attributes
 
@@ -137,6 +138,32 @@ class WeaviateVector(BaseVector):
 
         return client
 
+    def _ensure_connected(self) -> None:
+        """
+        Ensures the Weaviate client connection is active.
+
+        The Weaviate client connection can time out during idle periods (e.g. hours
+        between knowledge base creation and recall testing). This method checks the
+        connection state and reconnects if necessary, preventing
+        "WeaviateClient is closed" errors.
+        """
+        if self._client.is_connected():
+            return
+
+        try:
+            self._client.connect()
+            logger.info("Reconnected to Weaviate (collection=%s)", self._collection_name)
+        except Exception:
+            logger.warning(
+                "Failed to reconnect existing Weaviate client, creating a new one (collection=%s)",
+                self._collection_name,
+            )
+            try:
+                self._client.close()
+            except Exception:
+                pass
+            self._client = self._init_client(self._config)
+
     def get_type(self) -> str:
         """Returns the vector database type identifier."""
         return VectorType.WEAVIATE
@@ -173,6 +200,7 @@ class WeaviateVector(BaseVector):
 
         Uses Redis locking to prevent concurrent creation attempts.
         """
+        self._ensure_connected()
         lock_name = f"vector_indexing_lock_{self._collection_name}"
         with redis_client.lock(lock_name, timeout=20):
             cache_key = f"vector_indexing_{self._collection_name}"
@@ -255,6 +283,7 @@ class WeaviateVector(BaseVector):
 
         Batches insertions for efficiency and returns the list of inserted object IDs.
         """
+        self._ensure_connected()
         uuids = self._get_uuids(documents)
         texts = [d.page_content for d in documents]
         metadatas = [d.metadata for d in documents]
@@ -301,6 +330,7 @@ class WeaviateVector(BaseVector):
 
     def delete_by_metadata_field(self, key: str, value: str) -> None:
         """Deletes all objects matching a specific metadata field value."""
+        self._ensure_connected()
         if not self._client.collections.exists(self._collection_name):
             return
 
@@ -309,11 +339,13 @@ class WeaviateVector(BaseVector):
 
     def delete(self):
         """Deletes the entire collection from Weaviate."""
+        self._ensure_connected()
         if self._client.collections.exists(self._collection_name):
             self._client.collections.delete(self._collection_name)
 
     def text_exists(self, id: str) -> bool:
         """Checks if a document with the given doc_id exists in the collection."""
+        self._ensure_connected()
         if not self._client.collections.exists(self._collection_name):
             return False
 
@@ -332,6 +364,7 @@ class WeaviateVector(BaseVector):
 
         Silently ignores 404 errors for non-existent IDs.
         """
+        self._ensure_connected()
         if not self._client.collections.exists(self._collection_name):
             return
 
@@ -351,6 +384,7 @@ class WeaviateVector(BaseVector):
         Filters by document IDs if provided and applies score threshold.
         Returns documents sorted by relevance score.
         """
+        self._ensure_connected()
         if not self._client.collections.exists(self._collection_name):
             return []
 
@@ -398,6 +432,7 @@ class WeaviateVector(BaseVector):
 
         Filters by document IDs if provided and returns matching documents with vectors.
         """
+        self._ensure_connected()
         if not self._client.collections.exists(self._collection_name):
             return []