Feat: add chat complete api for no-stream and rewrite chat func for moscore (#253)

* feat: add chat complete

* feat: fix chat bug

Author: fridayL

src/memos/api/product_models.py

@@ -87,6 +87,18 @@ class ChatRequest(BaseRequest):
     moscube: bool = Field(False, description="Whether to use MemOSCube")
 
 
+class ChatCompleteRequest(BaseRequest):
+    """Request model for chat operations."""
+
+    user_id: str = Field(..., description="User ID")
+    query: str = Field(..., description="Chat query message")
+    mem_cube_id: str | None = Field(None, description="Cube ID to use for chat")
+    history: list[MessageDict] | None = Field(None, description="Chat history")
+    internet_search: bool = Field(False, description="Whether to use internet search")
+    moscube: bool = Field(False, description="Whether to use MemOSCube")
+    base_prompt: str | None = Field(None, description="Base prompt to use for chat")
+
+
 class UserCreate(BaseRequest):
     user_name: str | None = Field(None, description="Name of the user")
     role: str = Field("USER", description="Role of the user")

src/memos/api/routers/product_router.py

@@ -12,6 +12,7 @@
 from memos.api.context.dependencies import G, get_g_object
 from memos.api.product_models import (
     BaseResponse,
+    ChatCompleteRequest,
     ChatRequest,
     GetMemoryRequest,
     MemoryCreateRequest,
@@ -276,6 +277,33 @@ def generate_chat_response():
         raise HTTPException(status_code=500, detail=str(traceback.format_exc())) from err
 
 
+@router.post("/chat/complete", summary="Chat with MemOS (Complete Response)")
+def chat_complete(chat_req: ChatCompleteRequest):
+    """Chat with MemOS for a specific user. Returns complete response (non-streaming)."""
+    try:
+        mos_product = get_mos_product_instance()
+
+        # Collect all responses from the generator
+        content = mos_product.chat(
+            query=chat_req.query,
+            user_id=chat_req.user_id,
+            cube_id=chat_req.mem_cube_id,
+            history=chat_req.history,
+            internet_search=chat_req.internet_search,
+            moscube=chat_req.moscube,
+            base_prompt=chat_req.base_prompt,
+        )
+
+        # Return the complete response
+        return {"message": "Chat completed successfully", "data": {"response": content}}
+
+    except ValueError as err:
+        raise HTTPException(status_code=404, detail=str(traceback.format_exc())) from err
+    except Exception as err:
+        logger.error(f"Failed to start chat: {traceback.format_exc()}")
+        raise HTTPException(status_code=500, detail=str(traceback.format_exc())) from err
+
+
 @router.get("/users", summary="List all users", response_model=BaseResponse[list])
 def list_users():
     """List all registered users."""

src/memos/mem_os/product.py

@@ -709,6 +709,41 @@ def get_suggestion_query(
         response_json = json.loads(clean_response)
         return response_json["query"]
 
+    def chat(
+        self,
+        query: str,
+        user_id: str,
+        cube_id: str | None = None,
+        history: MessageList | None = None,
+        base_prompt: str | None = None,
+        internet_search: bool = False,
+        moscube: bool = False,
+        top_k: int = 10,
+    ) -> str:
+        """
+        Chat with LLM with memory references and complete response.
+        """
+        self._load_user_cubes(user_id, self.default_cube_config)
+        memories_result = super().search(
+            query,
+            user_id,
+            install_cube_ids=[cube_id] if cube_id else None,
+            top_k=top_k,
+            mode="fine",
+            internet_search=internet_search,
+            moscube=moscube,
+        )["text_mem"]
+        if memories_result:
+            memories_list = memories_result[0]["memories"]
+            memories_list = self._filter_memories_by_threshold(memories_list)
+        system_prompt = super()._build_system_prompt(memories_list, base_prompt)
+        current_messages = [
+            {"role": "system", "content": system_prompt},
+            {"role": "user", "content": query},
+        ]
+        response = self.chat_llm.generate(current_messages)
+        return response
+
     def chat_with_references(
         self,
         query: str,