From e76f77bcb79dbd471c2c0761e6c16202c3f329ee Mon Sep 17 00:00:00 2001
From: Timothy Jaeryang Baek <tim@openwebui.com>
Date: Sun, 9 Nov 2025 21:16:34 -0500
Subject: [PATCH] refac: stream chunk max buffer size

---
 backend/open_webui/env.py            | 10 ++++++----
 backend/open_webui/routers/openai.py |  4 ++--
 backend/open_webui/utils/misc.py     | 12 ++++++------
 3 files changed, 14 insertions(+), 12 deletions(-)

diff --git a/backend/open_webui/env.py b/backend/open_webui/env.py
index 4839396b21..b85316d955 100644
--- a/backend/open_webui/env.py
+++ b/backend/open_webui/env.py
@@ -570,16 +570,18 @@ else:
 
 
 CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = os.environ.get(
-    "CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE", "10485760" # 10MB
+    "CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE", ""
 )
 
 if CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE == "":
-    CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = 1024 * 1024 * 10
+    CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = None
 else:
     try:
-        CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = int(CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE)
+        CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = int(
+            CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE
+        )
     except Exception:
-        CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = 1024 * 1024 * 10
+        CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE = None
 
 
 ####################################
diff --git a/backend/open_webui/routers/openai.py b/backend/open_webui/routers/openai.py
index d3ec158cd9..b0f283c82d 100644
--- a/backend/open_webui/routers/openai.py
+++ b/backend/open_webui/routers/openai.py
@@ -45,7 +45,7 @@ from open_webui.utils.payload import (
 )
 from open_webui.utils.misc import (
     convert_logit_bias_input_to_json,
-    handle_large_stream_chunks,
+    stream_chunks_handler,
 )
 
 from open_webui.utils.auth import get_admin_user, get_verified_user
@@ -953,7 +953,7 @@ async def generate_chat_completion(
         if "text/event-stream" in r.headers.get("Content-Type", ""):
             streaming = True
             return StreamingResponse(
-                handle_large_stream_chunks(r.content),
+                stream_chunks_handler(r.content),
                 status_code=r.status,
                 headers=dict(r.headers),
                 background=BackgroundTask(
diff --git a/backend/open_webui/utils/misc.py b/backend/open_webui/utils/misc.py
index 49465fb3ea..ce16691365 100644
--- a/backend/open_webui/utils/misc.py
+++ b/backend/open_webui/utils/misc.py
@@ -542,21 +542,21 @@ def extract_urls(text: str) -> list[str]:
     return url_pattern.findall(text)
 
 
-def handle_large_stream_chunks(stream: aiohttp.StreamReader, max_buffer_size: int = CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE):
+def stream_chunks_handler(stream: aiohttp.StreamReader):
     """
     Handle stream response chunks, supporting large data chunks that exceed the original 16kb limit.
     When a single line exceeds max_buffer_size, returns an empty JSON string {} and skips subsequent data
     until encountering normally sized data.
 
     :param stream: The stream reader to handle.
-    :param max_buffer_size: The maximum buffer size in bytes, -1 means not handle large chunks, default is 10MB.
     :return: An async generator that yields the stream data.
     """
 
-    if max_buffer_size <= 0:
+    max_buffer_size = CHAT_STREAM_RESPONSE_CHUNK_MAX_BUFFER_SIZE
+    if max_buffer_size is None or max_buffer_size <= 0:
         return stream
 
-    async def handle_stream_chunks():
+    async def yield_safe_stream_chunks():
         buffer = b""
         skip_mode = False
 
@@ -603,5 +603,5 @@ def handle_large_stream_chunks(stream: aiohttp.StreamReader, max_buffer_size: in
         # Process remaining buffer data
         if buffer and not skip_mode:
             yield buffer
-    
-    return handle_stream_chunks()
+
+    return yield_safe_stream_chunks()