update glm 4.7 to novita

backend_api.py

@@ -98,7 +98,7 @@ def get_cached_client(model_id: str, provider: str = "auto"):
 
 # Define models and languages here to avoid importing Gradio UI
 AVAILABLE_MODELS = [
-    {"name": "GLM-4.7", "id": "zai-org/GLM-4.7", "description": "GLM-4.7 - Latest GLM model via HuggingFace (Default)", "supports_images": False},
+    {"name": "GLM-4.7", "id": "zai-org/GLM-4.7", "description": "GLM-4.7 - Latest GLM model via HuggingFace Router with Novita provider (Default)", "supports_images": False},
     {"name": "GLM-4.6", "id": "zai-org/GLM-4.6", "description": "GLM-4.6 model via HuggingFace with Cerebras provider", "supports_images": False},
     {"name": "GLM-4.6V 👁️", "id": "zai-org/GLM-4.6V:zai-org", "description": "GLM-4.6V vision model - supports image uploads for visual understanding", "supports_images": True},
     {"name": "DeepSeek V3", "id": "deepseek-ai/DeepSeek-V3", "description": "DeepSeek V3 - Fast model for code generation via HuggingFace Router with Novita provider", "supports_images": False},

backend_models.py

@@ -96,8 +96,8 @@ def get_real_model_id(model_id: str) -> str:
         return "zai-org/GLM-4.5:fireworks-ai"
     
     elif model_id == "zai-org/GLM-4.7":
-        # GLM-4.7 needs fastest provider suffix
-        return "zai-org/GLM-4.7:fastest"
+        # GLM-4.7 needs novita provider suffix
+        return "zai-org/GLM-4.7:novita"
     
     return model_id