diff --git a/studio/backend/routes/inference.py b/studio/backend/routes/inference.py
index a71fd1b8cc..8107e2d101 100644
--- a/studio/backend/routes/inference.py
+++ b/studio/backend/routes/inference.py
@@ -1137,7 +1137,7 @@ async def openai_chat_completions(
                     # Pre-compute static JSON envelope for content tokens (hot path).
                     # Must match model_dump_json(exclude_none=True): no finish_reason when null.
                     _chunk_prefix = f'{{"id":"{completion_id}","object":"chat.completion.chunk","created":{created},"model":{json.dumps(model_name)},"choices":[{{"index":0,"delta":{{"content":'
-                    _chunk_suffix = '}}]}'
+                    _chunk_suffix = "}}]}"
 
                     gen = gguf_generate_with_tools()
                     prev_text = ""
@@ -1275,7 +1275,7 @@ async def openai_chat_completions(
                     # Pre-compute static JSON envelope for content tokens (hot path).
                     # Must match model_dump_json(exclude_none=True): no finish_reason when null.
                     _chunk_prefix = f'{{"id":"{completion_id}","object":"chat.completion.chunk","created":{created},"model":{json.dumps(model_name)},"choices":[{{"index":0,"delta":{{"content":'
-                    _chunk_suffix = '}}]}'
+                    _chunk_suffix = "}}]}"
 
                     gen = gguf_generate()
                     prev_text = ""
@@ -1474,7 +1474,7 @@ async def openai_chat_completions(
                 # Pre-compute static JSON envelope for content tokens (hot path).
                 # Must match model_dump_json(exclude_none=True): no finish_reason when null.
                 _chunk_prefix = f'{{"id":"{completion_id}","object":"chat.completion.chunk","created":{created},"model":{json.dumps(model_name)},"choices":[{{"index":0,"delta":{{"content":'
-                _chunk_suffix = '}}]}'
+                _chunk_suffix = "}}]}"
 
                 gen = generate()
                 _disconnect_check_interval = 20