diff --git a/.clang-tidy.hash b/.clang-tidy.hash
index b448915e27..75f0fc9c98 100644
--- a/.clang-tidy.hash
+++ b/.clang-tidy.hash
@@ -1 +1 @@
-d15ae81646ac0ee76b2586716fe697f187281523ee6db566aed26542a9f98d1a
+a34e66dce18eaf9d2d094fedf6b9799bbd9ec1e734f76a29544e3f6d61786c3e
diff --git a/esphome/components/audio/__init__.py b/esphome/components/audio/__init__.py
index 7b03e4b6a7..6c721652e1 100644
--- a/esphome/components/audio/__init__.py
+++ b/esphome/components/audio/__init__.py
@@ -1,4 +1,5 @@
 import esphome.codegen as cg
+from esphome.components.esp32 import add_idf_component
 import esphome.config_validation as cv
 from esphome.const import CONF_BITS_PER_SAMPLE, CONF_NUM_CHANNELS, CONF_SAMPLE_RATE
 import esphome.final_validate as fv
@@ -165,4 +166,7 @@ def final_validate_audio_schema(
 
 
 async def to_code(config):
-    cg.add_library("esphome/esp-audio-libs", "2.0.1")
+    add_idf_component(
+        name="esphome/esp-audio-libs",
+        ref="2.0.3",
+    )
diff --git a/esphome/components/audio/audio_decoder.cpp b/esphome/components/audio/audio_decoder.cpp
index d1ad571a52..8f514468c4 100644
--- a/esphome/components/audio/audio_decoder.cpp
+++ b/esphome/components/audio/audio_decoder.cpp
@@ -300,7 +300,7 @@ FileDecoderState AudioDecoder::decode_mp3_() {
 
   // Advance read pointer to match the offset for the syncword
   this->input_transfer_buffer_->decrease_buffer_length(offset);
-  uint8_t *buffer_start = this->input_transfer_buffer_->get_buffer_start();
+  const uint8_t *buffer_start = this->input_transfer_buffer_->get_buffer_start();
 
   buffer_length = (int) this->input_transfer_buffer_->available();
   int err = esp_audio_libs::helix_decoder::MP3Decode(this->mp3_decoder_, &buffer_start, &buffer_length,
diff --git a/esphome/idf_component.yml b/esphome/idf_component.yml
index e19c1a1c87..9be787d0cd 100644
--- a/esphome/idf_component.yml
+++ b/esphome/idf_component.yml
@@ -1,6 +1,8 @@
 dependencies:
   bblanchon/arduinojson:
     version: "7.4.2"
+  esphome/esp-audio-libs:
+    version: 2.0.3
   espressif/esp-tflite-micro:
     version: 1.3.3~1
   espressif/esp32-camera:
diff --git a/platformio.ini b/platformio.ini
index a54aa0964c..62de5aa0d7 100644
--- a/platformio.ini
+++ b/platformio.ini
@@ -154,7 +154,6 @@ lib_deps =
     makuna/NeoPixelBus@2.8.0             ; neopixelbus
     esphome/ESP32-audioI2S@2.3.0         ; i2s_audio
     droscy/esp_wireguard@0.4.2           ; wireguard
-    esphome/esp-audio-libs@2.0.1         ; audio
     kahrendt/ESPMicroSpeechFeatures@1.1.0  ; micro_wake_word
 
 build_flags =
@@ -178,7 +177,6 @@ lib_deps =
     ${common:idf.lib_deps}
     droscy/esp_wireguard@0.4.2              ; wireguard
     kahrendt/ESPMicroSpeechFeatures@1.1.0   ; micro_wake_word
-    esphome/esp-audio-libs@2.0.1            ; audio
 build_flags =
     ${common:idf.build_flags}
     -Wno-nonnull-compare