From 7241b4e9b505ea433563a88916cece2d062cd063 Mon Sep 17 00:00:00 2001
From: AlexsanderHamir <alexsanderhamirgomesbaptista@gmail.com>
Date: Mon, 17 Nov 2025 12:59:48 -0800
Subject: [PATCH] add: comment above optimization

For anybody that would change this value for whatever reason, the comment makes the tradeoff clear.
---
 litellm/proxy/proxy_server.py | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/litellm/proxy/proxy_server.py b/litellm/proxy/proxy_server.py
index 16988205530..da592c90720 100644
--- a/litellm/proxy/proxy_server.py
+++ b/litellm/proxy/proxy_server.py
@@ -5236,6 +5236,10 @@ async def moderations(
 async def _audio_speech_chunk_generator(
     _response: HttpxBinaryResponseContent,
 ) -> AsyncGenerator[bytes, None]:
+    # chunk_size has a big impact on latency, it can't be too small or too large
+    # too small: latency is high
+    # too large: latency is low, but memory usage is high
+    # 8192 is a good compromise
     _generator = await _response.aiter_bytes(chunk_size=8192)
     async for chunk in _generator:
         yield chunk