diff --git a/computility-run.yaml b/computility-run.yaml
index b13d04c1..f02fcda2 100644
--- a/computility-run.yaml
+++ b/computility-run.yaml
@@ -8,9 +8,9 @@ command:
     - --served-model-name
     - llm
     - --max-model-len
-    - '256000'
+    - '131072'
     - --gpu-memory-utilization
-    - '0.95'
+    - '0.90'
     - --trust-remote-code
     - -tp
     - '4'
@@ -19,7 +19,7 @@ command:
     - --disable-log-requests
     - --disable-frontend-multiprocessing
     - --max-num-batched-tokens
-    - '4096'
+    - '8192'
     - --enable-chunked-prefill
     - --max-seq-len-to-capture
     - '32768'
diff --git a/qwen3_6_scripts/chat_utils.py b/qwen3_6_scripts/chat_utils.py
index 37d99629..9be41e0b 100644
--- a/qwen3_6_scripts/chat_utils.py
+++ b/qwen3_6_scripts/chat_utils.py
@@ -172,7 +172,8 @@ class BaseMultiModalItemTracker(ABC, Generic[_T]):
                 return "<image>"
             if model_type == "mllama":
                 return "<|image|>"
-            if model_type in ("qwen2_vl","qwen2_5_vl"):
+            if model_type in ("qwen2_vl", "qwen2_5_vl",
+                              "qwen3_5", "qwen3_5_moe"):
                 return "<|vision_start|><|image_pad|><|vision_end|>"
             if model_type == "molmo":
                 return ""
@@ -183,7 +184,8 @@ class BaseMultiModalItemTracker(ABC, Generic[_T]):
                 return "<|reserved_special_token_0|>"
             raise TypeError(f"Unknown model type: {model_type}")
         elif modality == "video":
-            if model_type in ("qwen2_vl","qwen2_5_vl"):
+            if model_type in ("qwen2_vl", "qwen2_5_vl",
+                              "qwen3_5", "qwen3_5_moe"):
                 return "<|vision_start|><|video_pad|><|vision_end|>"
             raise TypeError(f"Unknown model type: {model_type}")
         else:
diff --git a/qwen3_6_scripts/protocol.py b/qwen3_6_scripts/protocol.py
index 98792004..75d7ef15 100644
--- a/qwen3_6_scripts/protocol.py
+++ b/qwen3_6_scripts/protocol.py
@@ -408,6 +408,17 @@ class ChatCompletionRequest(OpenAIBaseModel):
         if data.get("max_completion_tokens") is not None and data.get("max_tokens") is None:
             data["max_tokens"] = data["max_completion_tokens"]
 
+        # Map thinking={enable:true/false} → chat_template_kwargs.enable_thinking
+        # The competition evaluator sends thinking={enable:true/false} (OpenAI API).
+        # Qwen3's chat template expects enable_thinking=True/False in kwargs.
+        thinking = data.get("thinking")
+        if isinstance(thinking, dict):
+            enable = thinking.get("enable")
+            if enable is not None:
+                ctk = data.get("chat_template_kwargs") or {}
+                ctk["enable_thinking"] = bool(enable)
+                data["chat_template_kwargs"] = ctk
+
         messages = data.get("messages")
         if not isinstance(messages, list):
             return data