diff --git a/homelab-ai-bot/voice.py b/homelab-ai-bot/voice.py
index e2e219b8..74d1f095 100644
--- a/homelab-ai-bot/voice.py
+++ b/homelab-ai-bot/voice.py
@@ -15,8 +15,9 @@ log = logging.getLogger("voice")
 
 _api_key: Optional[str] = None
 
-TTS_MODEL = "tts-1"
+TTS_MODEL = "gpt-4o-mini-tts"
 TTS_VOICE = "onyx"
+TTS_INSTRUCTIONS = "Sprich klar und deutlich auf Deutsch mit natuerlichem deutschem Akzent. Sachlicher, ruhiger Ton."
 WHISPER_MODEL = "whisper-1"
 
 
@@ -64,7 +65,13 @@ def synthesize(text: str) -> Optional[bytes]:
         r = requests.post(
             "https://api.openai.com/v1/audio/speech",
             headers={"Authorization": f"Bearer {key}", "Content-Type": "application/json"},
-            json={"model": TTS_MODEL, "input": text, "voice": TTS_VOICE, "response_format": "opus"},
+            json={
+                "model": TTS_MODEL,
+                "input": text,
+                "voice": TTS_VOICE,
+                "instructions": TTS_INSTRUCTIONS,
+                "response_format": "opus",
+            },
             timeout=30,
         )
         if r.ok: