diff --git a/firmware/esphome/packages/sendspin-addon-voice-assist.yaml b/firmware/esphome/packages/sendspin-addon-voice-assist.yaml
index d1010ed..8324b52 100644
--- a/firmware/esphome/packages/sendspin-addon-voice-assist.yaml
+++ b/firmware/esphome/packages/sendspin-addon-voice-assist.yaml
@@ -10,6 +10,9 @@
 
 media_player:
   - id: !extend external_media_player
+    announcement_pipeline:
+      sources:
+      - announcement_audio_file_source
     media_pipeline:
       sources:
       - audio_file_source
@@ -29,11 +32,11 @@ media_player:
               decibel_reduction: 0
               duration: 1.0s
 
-
-
 media_source:
   - platform: audio_file
     id: audio_file_source
+  - platform: audio_file
+    id: announcement_audio_file_source
 
 audio_file:
   - id: timer_finished_sound
diff --git a/firmware/esphome/packages/sendspin.yaml b/firmware/esphome/packages/sendspin.yaml
index e4f348b..b71e8cf 100644
--- a/firmware/esphome/packages/sendspin.yaml
+++ b/firmware/esphome/packages/sendspin.yaml
@@ -5,12 +5,17 @@
 # Requires substitutions:
 #   - task_stack_in_psram
 #   - friendly_name
+# Optional substitutions:
+#   - media_player_led_control: "true" (default) — set to "false"
+#     when voice-assist.yaml is also included, to avoid conflicts
+#     with voice-assist LED state management.
 # Requires i2s_audio to be configured in main config
 # !! Requires ESPHome 2026.5.0 or later for latest sendspin features and optimizations !!
 # !! at the moment this would only build with ESPHome (Dev) but should be available in ESPHome (main) by the mid-May !!
 # ============================================================
 
 substitutions:
+  media_player_led_control: "true"
   media_source_buffer_size: 200000
 
 globals:
@@ -86,12 +91,16 @@ media_player:
             - lambda: |-
                 id(announcement_triggered) = true;
                 id(current_volume) = id(external_media_player).volume;
-            - light.turn_on:
-                id: rgb_front_led
-                blue: 100%
-                red: 0%
-                green: 100%
-                brightness: 50%
+            - if:
+                condition:
+                  lambda: return ${media_player_led_control};
+                then:
+                  - light.turn_on:
+                      id: rgb_front_led
+                      blue: 100%
+                      red: 0%
+                      green: 100%
+                      brightness: 50%
             - mixer_speaker.apply_ducking:
                   id: media_mixer_input
                   decibel_reduction: 40
@@ -112,8 +121,12 @@ media_player:
             - media_player.volume_set:
                 id: external_media_player
                 volume: !lambda "return id(current_volume);"
-            - light.turn_off:
-                id: rgb_front_led
+            - if:
+                condition:
+                  lambda: return ${media_player_led_control};
+                then:
+                  - light.turn_off:
+                      id: rgb_front_led
             - lambda: id(announcement_triggered) = false;
 
     on_state:
diff --git a/firmware/esphome/packages/voice-assist-sendspin.yaml b/firmware/esphome/packages/voice-assist-sendspin.yaml
index 55e80b5..afe50ba 100644
--- a/firmware/esphome/packages/voice-assist-sendspin.yaml
+++ b/firmware/esphome/packages/voice-assist-sendspin.yaml
@@ -1,6 +1,7 @@
 substitutions:
   # Voice Assistant Configuration
   wake_word: okay_nabu
+  media_player_led_control: "false"  # disable media-player LED control; voice-assist manages LEDs
   speaker_volume: 60%
   voice_assist_idle_phase_id: '1'
   voice_assist_listening_phase_id: '2'