From b1f8f1aa070b8cbb9230ce10066b7819e25d36cf Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=9D=8E=E6=B3=B3=E6=A1=A6?=
 <39643373+liyonghua0910@users.noreply.github.com>
Date: Mon, 25 Aug 2025 18:31:57 +0800
Subject: [PATCH] [fix] fix completion stream api output_tokens not in usage
 (#3588)

---
 fastdeploy/entrypoints/openai/serving_completion.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/fastdeploy/entrypoints/openai/serving_completion.py b/fastdeploy/entrypoints/openai/serving_completion.py
index 8b938a388..252100e68 100644
--- a/fastdeploy/entrypoints/openai/serving_completion.py
+++ b/fastdeploy/entrypoints/openai/serving_completion.py
@@ -413,6 +413,7 @@ class OpenAIServingCompletion:
                                 usage=UsageInfo(
                                     prompt_tokens=len(prompt_batched_token_ids[idx]),
                                     completion_tokens=output_tokens[idx],
+                                    total_tokens=len(prompt_batched_token_ids[idx]) + output_tokens[idx],
                                 ),
                             )
                             yield f"data: {usage_chunk.model_dump_json(exclude_unset=True)}\n\n"