From fb1d51b93b82a8c7ee7a855cebd13cd8b024eb42 Mon Sep 17 00:00:00 2001
From: Abiba <abiba@sysloggh.com>
Date: Wed, 27 May 2026 07:22:30 +0000
Subject: [PATCH] restructure: routing prioritized by reasoning requirements
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Tier 1 (Lightweight): VLM → Dense → MoE     ≤500 tok, 1 turn
Tier 2 (Simple):      VLM → Dense → MoE     ≤15K tok, ≤12 turns (was 10K/10)
Tier 3 (Medium):      Dense → VLM → MoE     ≤25K tok
Tier 4 (Heavy):       MoE → Dense → VLM     >25K tok (MoE PRIMARY workhorse)
Tier 5 (Default):     MoE → Dense → VLM     MoE primary fallback

Target: MoE ~50% (heavy primary), VLM ~25% (raised simple + fallback),
        Dense ~25% (medium primary + heavy fallback)

Removed turn limit from Medium tier — Simple tier handles conversational
requests up to 12 turns now.
---
 router/router.py | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/router/router.py b/router/router.py
index 799c083..c34c751 100644
--- a/router/router.py
+++ b/router/router.py
@@ -218,8 +218,8 @@ def route(rd, tier):
         result = select_best_gpu(fallback, "lightweight_fallback")
         if result: return result
     
-    # TIER 2: Simple conversations — VLM primary (up to 10K tok), any prompt → VLM first, Dense second
-    if t <= 10000 and turns <= 10 and "qwen3.5-9b-vlm" in avail:
+    # TIER 2: Simple conversations — VLM primary (up to 15K tok), fastest for moderate chat
+    if t <= 15000 and turns <= 12 and "qwen3.5-9b-vlm" in avail:
         if not is_gpu_busy("qwen3.5-9b-vlm"):
             return {"model":"qwen3.5-9b-vlm","reason":"simple_conv"}
         # VLM busy — fall back to Dense, then MoE
@@ -227,20 +227,20 @@ def route(rd, tier):
         result = select_best_gpu(fallback, "simple_conv_fallback")
         if result: return result
     
-    # TIER 3: Medium complexity — Dense primary (speed), MoE fallback
-    if t <= 25000 and turns <= 15:
+    # TIER 3: Medium complexity — Dense primary, VLM fallback (quality + speed balance)
+    if t <= 25000:
         candidates = [m for m in ["qwen3.6-27B-code","qwen3.5-9b-vlm","qwen3.6-35B-A3B"] if m in avail]
         result = select_best_gpu(candidates, "medium")
         if result: return result
     
-    # TIER 4: Heavy reasoning — Dense first, VLM second (262K ctx), MoE last (thermal)
-    if t > 25000 or turns > 15:
-        candidates = [m for m in ["qwen3.6-27B-code","qwen3.5-9b-vlm","qwen3.6-35B-A3B"] if m in avail]
+    # TIER 4: Heavy reasoning — MoE primary (workhorse), Dense fallback
+    if t > 25000:
+        candidates = [m for m in ["qwen3.6-35B-A3B","qwen3.6-27B-code","qwen3.5-9b-vlm"] if m in avail]
         result = select_best_gpu(candidates, "heavy_reasoning")
         if result: return result
     
-    # TIER 5: Default — balanced distribution: Dense first (speed), MoE second (capacity)
-    candidates = [m for m in ["qwen3.6-27B-code","qwen3.5-9b-vlm","qwen3.6-35B-A3B"] if m in avail]
+    # TIER 5: Default — MoE primary, Dense fallback
+    candidates = [m for m in ["qwen3.6-35B-A3B","qwen3.6-27B-code","qwen3.5-9b-vlm"] if m in avail]
     result = select_best_gpu(candidates, "default")
     if result: return result
     return {"model":avail[0],"reason":"last_resort"}