fix: swap to gpt-realtime-whisper for STT

Replaces gpt-4o-mini-transcribe (/usr/bin/bash.003/min) with gpt-realtime-whisper
(/usr/bin/bash.017/min). Expected to reduce transcription latency from ~2.6s
to ~1s due to the model's realtime optimization.

backend/main.py

@@ -85,7 +85,7 @@ async def transcribe_audio(audio_bytes: bytes) -> str | None:
     client = get_openai()
     try:
         transcript = await client.audio.transcriptions.create(
-            model="gpt-4o-mini-transcribe",
+            model="gpt-realtime-whisper",
             file=("audio.webm", audio_bytes, "audio/webm"),
             response_format="text",
         )