Upload hf_inference.py

core/hf_inference.py

@@ -333,9 +333,19 @@ class HFInferenceClient:
         if not self.token:
             raise ValueError("HF_TOKEN is required for inference")
         
-        # Create clients
-        self.sync_client = InferenceClient(token=self.token)
-        self.async_client = AsyncInferenceClient(token=self.token)
+        # New HuggingFace router URL (api-inference is deprecated)
+        # The huggingface_hub library should handle this automatically with newer versions,
+        # but we set it explicitly for compatibility
+        router_url = "https://router.huggingface.co"
+        
+        # Create clients with explicit base_url for new router
+        try:
+            self.sync_client = InferenceClient(token=self.token, base_url=router_url)
+            self.async_client = AsyncInferenceClient(token=self.token, base_url=router_url)
+        except TypeError:
+            # Fallback for older huggingface_hub versions that don't support base_url
+            self.sync_client = InferenceClient(token=self.token)
+            self.async_client = AsyncInferenceClient(token=self.token)
     
     def get_model_for_agent(self, agent_name: str) -> str:
         """Get the appropriate model for an agent."""