fix(nvidia): timeout 60s→30s，NVIDIA 第一保免費，失敗轉 Gemini

- nvidia_provider.py: NVIDIA_TIMEOUT 60→30s
- models.json: timeout_seconds 60→30s
- configmap: NEMOTRON_TIMEOUT_SECONDS 45→30s, fallback 恢復 nvidia 第一
目標: Nemo 有足夠時間回應(free)，失敗快速轉 Gemini(備援)，整體機制可運作

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

apps/api/models.json

@@ -119,11 +119,11 @@
         "temperature": 0.0,
         "max_tokens": 1024
       },
-      "timeout_seconds": 60,
+      "timeout_seconds": 30,
       "cost": {
         "per_1k_tokens": 0,
         "currency": "USD",
-        "notes": "Free tier (2026-03-29)"
+        "notes": "Free tier (2026-03-29) - timeout 30s，失敗後轉 Gemini"
       },
       "auth": {
         "type": "header",

apps/api/src/services/nvidia_provider.py

@@ -117,8 +117,9 @@ NVIDIA_API_URL = "https://integrate.api.nvidia.com/v1/chat/completions"
 # 預設模型 (2026-03-31 ogt: 恢復為 nemotron-mini-4b-instruct)
 NVIDIA_DEFAULT_MODEL = "nvidia/nemotron-mini-4b-instruct"
 
-# 請求超時 (秒) - Nemotron 延遲 11-45s
-NVIDIA_TIMEOUT = 60.0
+# 請求超時 (秒) - 2026-04-01 ogt: 30s 讓 Nemo 有機會回應，失敗後轉 Gemini
+# 原 60s 太長；15s 太短 (Nemo 有時需 20-40s)；30s 是實用平衡點
+NVIDIA_TIMEOUT = 30.0
 
 # 重試次數
 MAX_RETRIES = 2

k8s/awoooi-prod/04-configmap.yaml

@@ -35,9 +35,9 @@ data:
   CORS_ORIGINS: '["https://awoooi.wooo.work","http://localhost:3000","http://localhost:3001"]'
 
   # AI 配置 (JSON array 格式 for pydantic-settings)
-  # 2026-04-01 ogt: Gemini 優先 (快速，~100ms) → Ollama(本地備援) → NVIDIA(60s timeout，降為最後)
-  # NVIDIA 降序原因: free tier 固定 60s timeout，阻塞整條 fallback 鏈
-  AI_FALLBACK_ORDER: '["gemini","ollama","nvidia","claude"]'
+  # 2026-04-01 ogt: NVIDIA 第一 (免費) → Gemini(備援 ~100ms) → Ollama(本地) → Claude
+  # NVIDIA timeout 降為 15s，快速失敗後轉 Gemini，避免免費 tier 阻塞整條鏈
+  AI_FALLBACK_ORDER: '["nvidia","gemini","ollama","claude"]'
   AI_CACHE_TTL: "3600"
 
   # ============================================================================
@@ -45,7 +45,7 @@ data:
   # 2026-03-31 ogt: 統帥需求 - @tsenyangbot 對話 + Incident 雙軌
   # ============================================================================
   ENABLE_NEMOTRON_COLLABORATION: "true"
-  NEMOTRON_TIMEOUT_SECONDS: "45"
+  NEMOTRON_TIMEOUT_SECONDS: "30"
   NEMOTRON_ASYNC_UPDATE: "true"
 
   # ============================================================================