refac

Author: tjbck

backend/open_webui/routers/openai.py

@@ -858,6 +858,9 @@ def convert_to_responses_payload(payload: dict) -> dict:
     if 'max_tokens' in responses_payload:
         responses_payload['max_output_tokens'] = responses_payload.pop('max_tokens')
 
+    if 'max_completion_tokens' in responses_payload:
+        responses_payload['max_output_tokens'] = responses_payload.pop('max_completion_tokens')
+
     # Remove Chat Completions-only parameters not supported by the Responses API
     for unsupported_key in (
         'stream_options',
@@ -896,11 +899,36 @@ def convert_to_responses_payload(payload: dict) -> dict:
 
 def convert_responses_result(response: dict) -> dict:
     """
-    Convert non-streaming Responses API result.
-    Just add done flag - pass through raw response, frontend handles output.
+    Convert non-streaming Responses API result to Chat Completions format.
+
+    Extracts text from message output items so all downstream consumers
+    (frontend tasks, get_content_from_response) work without modification.
     """
-    response['done'] = True
-    return response
+    output_items = response.get('output', [])
+
+    content = ''
+    for item in output_items:
+        if item.get('type') == 'message':
+            for part in item.get('content', []):
+                if part.get('type') == 'output_text':
+                    content += part.get('text', '')
+
+    return {
+        'id': response.get('id', ''),
+        'object': 'chat.completion',
+        'model': response.get('model', ''),
+        'choices': [
+            {
+                'index': 0,
+                'message': {
+                    'role': 'assistant',
+                    'content': content,
+                },
+                'finish_reason': 'stop',
+            }
+        ],
+        'usage': response.get('usage', {}),
+    }
 
 
 @router.post('/chat/completions')

backend/open_webui/utils/middleware.py

@@ -3367,6 +3367,9 @@ def set_last_text(out, text):
             usage = None
             prior_output = []
 
+            def full_output():
+                return prior_output + output if prior_output else output
+
             reasoning_tags_param = metadata.get('params', {}).get('reasoning_tags')
             DETECT_REASONING_TAGS = reasoning_tags_param is not False
             DETECT_CODE_INTERPRETER = metadata.get('features', {}).get('code_interpreter', False)
@@ -3476,8 +3479,8 @@ async def flush_pending_delta_data(threshold: int = 0):
                                     output, response_metadata = handle_responses_streaming_event(data, output)
 
                                     processed_data = {
-                                        'output': prior_output + output,
-                                        'content': serialize_output(prior_output + output),
+                                        'output': full_output(),
+                                        'content': serialize_output(full_output()),
                                     }
 
                                     # print(data)
@@ -3832,13 +3835,13 @@ async def flush_pending_delta_data(threshold: int = 0):
                                                 metadata['chat_id'],
                                                 metadata['message_id'],
                                                 {
-                                                    'content': serialize_output(output),
-                                                    'output': output,
+                                                    'content': serialize_output(full_output()),
+                                                    'output': full_output(),
                                                 },
                                             )
                                         else:
                                             data = {
-                                                'content': serialize_output(output),
+                                                'content': serialize_output(full_output()),
                                             }
 
                                 if delta:
@@ -3952,26 +3955,32 @@ async def flush_pending_delta_data(threshold: int = 0):
                     response_tool_calls = tool_calls.pop(0)
 
                     # Append function_call items for each tool call
+                    # (Responses API already has them from streaming, so skip duplicates)
+                    existing_call_ids = {
+                        item.get('call_id') for item in output
+                        if item.get('type') == 'function_call'
+                    }
                     for tc in response_tool_calls:
                         call_id = tc.get('id', '')
-                        func = tc.get('function', {})
-                        output.append(
-                            {
-                                'type': 'function_call',
-                                'id': call_id or output_id('fc'),
-                                'call_id': call_id,
-                                'name': func.get('name', ''),
-                                'arguments': func.get('arguments', '{}'),
-                                'status': 'in_progress',
-                            }
-                        )
+                        if call_id not in existing_call_ids:
+                            func = tc.get('function', {})
+                            output.append(
+                                {
+                                    'type': 'function_call',
+                                    'id': call_id or output_id('fc'),
+                                    'call_id': call_id,
+                                    'name': func.get('name', ''),
+                                    'arguments': func.get('arguments', '{}'),
+                                    'status': 'in_progress',
+                                }
+                            )
 
                     await event_emitter(
                         {
                             'type': 'chat:completion',
                             'data': {
-                                'content': serialize_output(output),
-                                'output': output,
+                                'content': serialize_output(full_output()),
+                                'output': full_output(),
                             },
                         }
                     )