Update app.py

app.py

@@ -6,8 +6,9 @@ from fastapi.responses import JSONResponse, HTMLResponse
 app = FastAPI()
 
 HF_TOKEN = os.environ.get("HF_TOKEN")
-# The standard address for a text-generation model
-MODEL_URL = "https://api-inference.huggingface.co/models/danosethrus/EthioDoc"
+
+# This is the modern 'Router' URL that is more reliable
+MODEL_URL = "https://router.huggingface.co/hf-inference/models/danosethrus/EthioDoc"
 
 @app.get("/", response_class=HTMLResponse)
 async def home():
@@ -19,7 +20,7 @@ async def ask_ai(request: Request):
     data = await request.json()
     headers = {"Authorization": f"Bearer {HF_TOKEN}"}
     
-    # We add 'wait_for_model' so it doesn't error out while loading the 4GB file
+    # We send the question
     payload = {
         "inputs": data.get("inputs", ""),
         "options": {"wait_for_model": True}
@@ -27,8 +28,9 @@ async def ask_ai(request: Request):
 
     response = requests.post(MODEL_URL, headers=headers, json=payload)
     
-    # This will show the real answer or error in your Logs
-    print(f"DEBUG Status: {response.status_code}")
-    print(f"DEBUG Response: {response.text}")
+    # This checks if it's still 404 before trying to read JSON (to avoid the crash)
+    if response.status_code != 200:
+        print(f"FAILED: {response.status_code} - {response.text}")
+        return JSONResponse({"error": "AI is still booting up or URL is wrong."}, status_code=response.status_code)
 
     return JSONResponse(content=response.json())