Update app.py

app.py

@@ -2,6 +2,7 @@ import os
 import re
 import json
 import io
+import time
 import traceback
 import contextlib
 import tempfile
@@ -12,8 +13,19 @@ import pandas as pd
 
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
-GROQ_MODEL = os.getenv("GROQ_MODEL", "llama-3.3-70b-versatile")
-MAX_TOOL_ITERATIONS = 8
+# Primary model first, then fallbacks used when rate-limited.
+# llama-3.1-8b-instant has much higher daily token quota on the free tier.
+GROQ_MODELS = [
+    m.strip()
+    for m in os.getenv(
+        "GROQ_MODELS",
+        "llama-3.3-70b-versatile,llama-3.1-8b-instant",
+    ).split(",")
+    if m.strip()
+]
+MAX_TOOL_ITERATIONS = 6
+TOOL_RESULT_MAX_CHARS = 3000
+ANSWER_CACHE_PATH = os.getenv("ANSWER_CACHE_PATH", "answers_cache.json")
 
 
 # ---------------------------------------------------------------------------