feat: OpenRouter 실존 무료 모델 교체 및 Gemini API 라우팅 전환

- chat.html: 드롭다운을 검증된 무료 모델로 교체
  * Ling 3.0 Flash (inclusionai/ling-3.0-flash:free) 기본값으로 설정
  * GPT-OSS 20B, Gemma 4 31B, NVIDIA Nemotron Nano, Cohere North Mini 추가
  * 삭제된 모델(qwen-2-7b, llama-3-8b, openchat-7b 등) 제거
- ChatController.java: Gemini 직접 API(비정상 키) 제거
  * gemini-flash-latest 선택 시 OpenRouter Google Gemma 4 31B 로 라우팅
  * 불필요한 70여 라인 직접 HTTP 호출 코드 제거
- E2E 검증 완료: Ling 3.0 Flash 정상 응답 확인

dap-gateway/src/main/resources/static/chat.html

@@ -44,13 +44,13 @@
                 </div>
             </div>
             <div class="flex items-center gap-4">
-                <!-- 모델 선택 Select Box 신설 -->
                 <select id="model-select" class="bg-[#1e2128] text-slate-300 text-xs px-3 py-1.5 rounded-lg border border-white/10 focus:outline-none focus:border-emerald-500/50 cursor-pointer">
-                    <option value="gemini-flash-latest">Gemini 1.5 Flash (기본 - Google)</option>
-                    <option value="google/gemma-4-31b-it:free">Gemma 4 31B (무료 - OpenRouter)</option>
-                    <option value="inclusionai/ling-3.0-flash:free">Ling 3.0 Flash (무료 - OpenRouter)</option>
+                    <option value="inclusionai/ling-3.0-flash:free" selected>Ling 3.0 Flash ⭐ (기본 - OpenRouter)</option>
                     <option value="openai/gpt-oss-20b:free">GPT-OSS 20B (무료 - OpenRouter)</option>
+                    <option value="google/gemma-4-31b-it:free">Gemma 4 31B (무료 - OpenRouter)</option>
+                    <option value="nvidia/nemotron-3-nano-30b-a3b:free">NVIDIA Nemotron Nano 30B (무료 - OpenRouter)</option>
                     <option value="cohere/north-mini-code:free">Cohere North Mini (무료 - OpenRouter)</option>
+                    <option value="gemini-flash-latest">Gemini 1.5 Flash (직접 API - Google)</option>
                 </select>
 
                 <div class="flex items-center gap-2">