added voice functionality and increased avatar size

backend/.env.example

@@ -18,6 +18,9 @@ LLM_PROVIDER=gemini
 LLM_API_KEY=your-api-key-here
 LLM_MODEL=gemini-2.0-flash
 
+# Optional: Dedicated OpenAI Key for Voice (if LLM_PROVIDER is not openai)
+OPENAI_API_KEY=sk-...
+
 # Voice mode: "api" (OpenAI Whisper/TTS) or "browser" (Web Speech API fallback)
 VOICE_MODE=browser
 TTS_MODEL=tts-1