fix: routing priority — MoE first, VLM second, Dense last (slow)

All tiers now follow MoE → VLM → Dense priority order since
Dense (RTX 3090) can be slow. VLM acts as overflow absorber.

router/router.py

@@ -168,7 +168,7 @@ def route(rd, tier):
         if not is_gpu_busy("qwen3.5-9b-vlm"):
             return {"model":"qwen3.5-9b-vlm","reason":"lightweight"}
         # VLM busy — fall back to Dense, then MoE
-        fallback = [m for m in ["qwen3.6-27B-code","qwen3.6-35B-A3B"] if m in avail]
+        fallback = [m for m in ["qwen3.6-35B-A3B","qwen3.6-27B-code"] if m in avail]
         result = select_best_gpu(fallback, "lightweight_fallback")
         if result: return result
     
@@ -182,15 +182,14 @@ def route(rd, tier):
     
     # TIER 3: Heavy reasoning — large context or very long conversations
     if t > 4000 or turns > 8:
-        candidates = [m for m in ["qwen3.6-27B-code","qwen3.6-35B-A3B","qwen3.5-9b-vlm"] if m in avail]
+        candidates = [m for m in ["qwen3.6-35B-A3B","qwen3.5-9b-vlm","qwen3.6-27B-code"] if m in avail]
         result = select_best_gpu(candidates, "heavy_reasoning")
         if result: return result
     
-    # TIER 4: Default — Dense preferred for medium tasks, MoE as workhorse, VLM as overflow
-    if turns <= 6 and t <= 4000:
-        # Medium complexity — try Dense first, then MoE, then VLM
-        candidates = [m for m in ["qwen3.6-27B-code","qwen3.6-35B-A3B","qwen3.5-9b-vlm"] if m in avail]
-        result = select_best_gpu(candidates, "medium_task")
+    # TIER 4: Default — MoE first, VLM helps, Dense last (slow)
+    if t <= 4000:
+        candidates = [m for m in ["qwen3.6-35B-A3B","qwen3.5-9b-vlm","qwen3.6-27B-code"] if m in avail]
+        result = select_best_gpu(candidates, "default")
         if result: return result
     
     # Fallback — best available