diff --git a/api/app/controllers/chunk_controller.py b/api/app/controllers/chunk_controller.py
index 620d8a1a..988aa706 100644
--- a/api/app/controllers/chunk_controller.py
+++ b/api/app/controllers/chunk_controller.py
@@ -441,14 +441,14 @@ async def retrieve_chunks(
     # 1 participle search, 2 semantic search, 3 hybrid search
     match retrieve_data.retrieve_type:
         case chunk_schema.RetrieveType.PARTICIPLE:
-            rs = vector_service.search_by_full_text(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.similarity_threshold)
+            rs = vector_service.search_by_full_text(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.similarity_threshold, file_names_filter=retrieve_data.file_names_filter)
             return success(data=rs, msg="retrieval successful")
         case chunk_schema.RetrieveType.SEMANTIC:
-            rs = vector_service.search_by_vector(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.vector_similarity_weight)
+            rs = vector_service.search_by_vector(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.vector_similarity_weight, file_names_filter=retrieve_data.file_names_filter)
             return success(data=rs, msg="retrieval successful")
         case _:
-            rs1 = vector_service.search_by_vector(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.vector_similarity_weight)
-            rs2 = vector_service.search_by_full_text(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.similarity_threshold)
+            rs1 = vector_service.search_by_vector(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.vector_similarity_weight, file_names_filter=retrieve_data.file_names_filter)
+            rs2 = vector_service.search_by_full_text(query=retrieve_data.query, top_k=retrieve_data.top_k, indices=indices, score_threshold=retrieve_data.similarity_threshold, file_names_filter=retrieve_data.file_names_filter)
             # Efficient deduplication
             seen_ids = set()
             unique_rs = []
diff --git a/api/app/schemas/chunk_schema.py b/api/app/schemas/chunk_schema.py
index cef9b9cb..ce8f70f2 100644
--- a/api/app/schemas/chunk_schema.py
+++ b/api/app/schemas/chunk_schema.py
@@ -46,6 +46,7 @@ class ChunkUpdate(BaseModel):
 class ChunkRetrieve(BaseModel):
     query: str
     kb_ids: list[uuid.UUID]
+    file_names_filter: list[str] | None = Field(None)
     similarity_threshold: float | None = Field(None)
     vector_similarity_weight: float | None = Field(None)
     top_k: int | None = Field(None)