refac: ollama response

backend/open_webui/utils/middleware.py

@@ -1866,9 +1866,11 @@ async def process_chat_response(
 
                                     value = delta.get("content")
 
-                                    reasoning_content = delta.get(
-                                        "reasoning_content"
-                                    ) or delta.get("reasoning")
+                                    reasoning_content = (
+                                        delta.get("reasoning_content")
+                                        or delta.get("reasoning")
+                                        or delta.get("thinking")
+                                    )
                                     if reasoning_content:
                                         if (
                                             not content_blocks

backend/open_webui/utils/misc.py

@@ -208,6 +208,7 @@ def openai_chat_message_template(model: str):
 def openai_chat_chunk_message_template(
     model: str,
     content: Optional[str] = None,
+    reasoning_content: Optional[str] = None,
     tool_calls: Optional[list[dict]] = None,
     usage: Optional[dict] = None,
 ) -> dict:
@@ -220,6 +221,9 @@ def openai_chat_chunk_message_template(
     if content:
         template["choices"][0]["delta"]["content"] = content
 
+    if reasoning_content:
+        template["choices"][0]["delta"]["reasonsing_content"] = reasoning_content
+
     if tool_calls:
         template["choices"][0]["delta"]["tool_calls"] = tool_calls
 
@@ -234,6 +238,7 @@ def openai_chat_chunk_message_template(
 def openai_chat_completion_message_template(
     model: str,
     message: Optional[str] = None,
+    reasoning_content: Optional[str] = None,
     tool_calls: Optional[list[dict]] = None,
     usage: Optional[dict] = None,
 ) -> dict:
@@ -241,8 +246,9 @@ def openai_chat_completion_message_template(
     template["object"] = "chat.completion"
     if message is not None:
         template["choices"][0]["message"] = {
-            "content": message,
             "role": "assistant",
+            "content": message,
+            **({"reasoning_content": reasoning_content} if reasoning_content else {}),
             **({"tool_calls": tool_calls} if tool_calls else {}),
         }
 

backend/open_webui/utils/response.py

@@ -105,6 +105,7 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)
 
         model = data.get("model", "ollama")
         message_content = data.get("message", {}).get("content", None)
+        reasoning_content = data.get("message", {}).get("thinking", None)
         tool_calls = data.get("message", {}).get("tool_calls", None)
         openai_tool_calls = None
 
@@ -118,7 +119,7 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)
             usage = convert_ollama_usage_to_openai(data)
 
         data = openai_chat_chunk_message_template(
-            model, message_content, openai_tool_calls, usage
+            model, message_content, reasoning_content, openai_tool_calls, usage
         )
 
         line = f"data: {json.dumps(data)}\n\n"