diff --git a/docker-compose.macmini.yml b/docker-compose.macmini.yml
index 4ebeca8e5..2b8fd5e30 100644
--- a/docker-compose.macmini.yml
+++ b/docker-compose.macmini.yml
@@ -980,7 +980,7 @@ services:
       MATRIX_ALLOWED_ROOMS: ${MATRIX_TTS_BOT_ROOMS:-}
       TTS_URL: http://host.docker.internal:3022
       TTS_API_KEY: ${TTS_INTERNAL_API_KEY:-}
-      DEFAULT_VOICE: af_heart
+      DEFAULT_VOICE: de_thorsten
       DEFAULT_SPEED: 1.0
       MAX_TEXT_LENGTH: 500
     volumes:
diff --git a/services/matrix-tts-bot/src/tts/tts.service.ts b/services/matrix-tts-bot/src/tts/tts.service.ts
index f9b79db38..f3348ee40 100644
--- a/services/matrix-tts-bot/src/tts/tts.service.ts
+++ b/services/matrix-tts-bot/src/tts/tts.service.ts
@@ -122,7 +122,7 @@ export class TtsService {
 	/**
 	 * Synthesize text to speech - auto-detects language
 	 */
-	async synthesize(text: string, voice: string = 'af_heart', speed: number = 1.0): Promise<Buffer> {
+	async synthesize(text: string, voice: string = 'de_thorsten', speed: number = 1.0): Promise<Buffer> {
 		// Auto-detect language if using English voice but text is German
 		const textIsGerman = this.isGerman(text);
 		const voiceIsGerman = this.isGermanVoice(voice);