From ea97cb1a78a623c56bc04bd5c0e17db19415ea7b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Aleix=20Conchillo=20Flaqu=C3=A9?= <aleix@daily.co>
Date: Thu, 14 May 2026 13:22:50 -0700
Subject: [PATCH] Add changelog for #4497

---
 changelog/4497.fixed.md | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 changelog/4497.fixed.md

diff --git a/changelog/4497.fixed.md b/changelog/4497.fixed.md
new file mode 100644
index 000000000..b9ba8b676
--- /dev/null
+++ b/changelog/4497.fixed.md
@@ -0,0 +1 @@
+- Fixed dropped audio in streaming TTS services whose wire protocol doesn't echo `context_id` back on incoming audio (Sarvam, Smallest, Soniox, Inworld, and others). Previously, audio that arrived between contexts or at the very start of a turn was tagged with `context_id=None` and silently dropped with an "unable to append audio to context: no context ID provided" debug log. `TTSService.get_active_audio_context_id()` now falls back to the synthesis-side `_turn_context_id` when the playback cursor isn't set yet.