feat: Reorder models by parameter count and move model dropdown in UI

- Models now ordered by parameter count (0.1B → 30B) ascending
- Moved model dropdown to top level (below Output Language, above Upload File)
- Model selection now more prominent and accessible
- Removed duplicate enable_reasoning checkbox from Advanced Settings
- Added thinking headroom calculation for max_tokens

app.py

@@ -190,20 +190,6 @@ AVAILABLE_MODELS = {
             "repeat_penalty": 1.0,
         },
     },
-    "qwen3_30b_thinking_q1": {
-        "name": "Qwen3 30B Thinking (256K Context)",
-        "repo_id": "unsloth/Qwen3-30B-A3B-Thinking-2507-GGUF",
-        "filename": "*TQ1_0.gguf",
-        "max_context": 262144,
-        "default_temperature": 0.6,
-        "supports_toggle": False,  # Thinking-only mode
-        "inference_settings": {
-            "temperature": 0.6,
-            "top_p": 0.95,
-            "top_k": 20,
-            "repeat_penalty": 1.0,
-        },
-    },
     "granite4_tiny_q3": {
         "name": "Granite 4.0 Tiny 7B (128K Context)",
         "repo_id": "unsloth/granite-4.0-h-tiny-GGUF",
@@ -246,6 +232,20 @@ AVAILABLE_MODELS = {
             "repeat_penalty": 1.05,
         },
     },
+    "qwen3_30b_thinking_q1": {
+        "name": "Qwen3 30B Thinking (256K Context)",
+        "repo_id": "unsloth/Qwen3-30B-A3B-Thinking-2507-GGUF",
+        "filename": "*TQ1_0.gguf",
+        "max_context": 262144,
+        "default_temperature": 0.6,
+        "supports_toggle": False,  # Thinking-only mode
+        "inference_settings": {
+            "temperature": 0.6,
+            "top_p": 0.95,
+            "top_k": 20,
+            "repeat_penalty": 1.0,
+        },
+    },
 }
 
 DEFAULT_MODEL_KEY = "qwen3_600m_q4"
@@ -1019,6 +1019,24 @@ def create_interface():
                         info="Choose the target language for your summary"
                     )
                     
+                    # Model Selection - Moved to top level for easy access
+                    gr.HTML('<div class="section-header" style="margin-top: 20px;"><span class="section-icon">🤖</span> Model</div>')
+                    
+                    model_dropdown = gr.Dropdown(
+                        choices=[(info["name"], key) for key, info in AVAILABLE_MODELS.items()],
+                        value=DEFAULT_MODEL_KEY,
+                        label="Select Model",
+                        info="Models ordered by size (0.6B to 30B). Smaller = faster. Large files need bigger context."
+                    )
+                    
+                    enable_reasoning = gr.Checkbox(
+                        value=True,
+                        label="Enable Reasoning Mode",
+                        info="Uses /think for deeper analysis (slower) or /no_think for direct output (faster). Only available for Qwen3 models.",
+                        interactive=True,
+                        visible=AVAILABLE_MODELS[DEFAULT_MODEL_KEY].get("supports_toggle", False)
+                    )
+                    
                     gr.HTML('<div class="section-header" style="margin-top: 20px;"><span class="section-icon">📤</span> Upload File</div>')
                     
                     file_input = gr.File(
@@ -1030,19 +1048,6 @@ def create_interface():
                     
                     with gr.Accordion("⚙️ Advanced Settings", open=False):
                         with gr.Group(elem_classes=["advanced-settings"]):
-                            model_dropdown = gr.Dropdown(
-                                choices=[(info["name"], key) for key, info in AVAILABLE_MODELS.items()],
-                                value=DEFAULT_MODEL_KEY,
-                                label="Model",
-                                info="Smaller = faster. Large files need models with bigger context."
-                            )
-                            enable_reasoning = gr.Checkbox(
-                                value=True,
-                                label="Enable Reasoning Mode",
-                                info="Uses /think for deeper analysis (slower) or /no_think for direct output (faster). Only available for Qwen3 models.",
-                                interactive=True,
-                                visible=AVAILABLE_MODELS[DEFAULT_MODEL_KEY].get("supports_toggle", False)
-                            )
                             temperature_slider = gr.Slider(
                                 minimum=0.0,
                                 maximum=2.0,