From 3fee91ddecb4eb8a35da4ca592d2eae771ed33d9 Mon Sep 17 00:00:00 2001
From: Paul Kompfner <paul@daily.co>
Date: Mon, 18 May 2026 12:06:48 -0400
Subject: [PATCH] Drop redundant changelog entry for OpenAI Realtime example
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The OpenAI Realtime story didn't add any service-level code — just a
new example. The original 4480.added.md entry already describes the
feature as "a realtime service like Gemini Live," which generalizes
to OpenAI Realtime.
---
 changelog/4480.added.2.md | 1 -
 1 file changed, 1 deletion(-)
 delete mode 100644 changelog/4480.added.2.md

diff --git a/changelog/4480.added.2.md b/changelog/4480.added.2.md
deleted file mode 100644
index 26417a9c3..000000000
--- a/changelog/4480.added.2.md
+++ /dev/null
@@ -1 +0,0 @@
-- The `wait_for_transcript_to_end_user_turn=False` pattern also works with OpenAI Realtime. Set `turn_detection=False` inside `OpenAIRealtimeLLMService.Settings.session_properties.audio.input` to disable OpenAI's server-side VAD; the service then drives turn boundaries from local turn detection, sending `input_audio_buffer.commit` + `response.create` on `UserStoppedSpeakingFrame`. See `examples/realtime/realtime-openai-local-vad.py` for the full pattern.