Fix: Groq model, error handling for all modes, LITE_MODE fallbacks

app/api.py

@@ -487,13 +487,18 @@ def deep_research(req: ChatRequest):
     memory.add(ws, "user", q)
 
     try:
-        state = deep_graph.run(q)
-        # state is a dict (TypedDict), not an object
-        answer = state.get("final_answer", "No answer generated.")
-        sources = state.get("sources", [])
+        if deep_graph is None:
+            # LITE_MODE fallback - use web search instead
+            state = web_graph.run(q)
+            answer = state.get("answer", "No answer generated.")
+            sources = state.get("sources", [])
+        else:
+            state = deep_graph.run(q)
+            answer = state.get("final_answer", "No answer generated.")
+            sources = state.get("sources", [])
     except Exception as e:
-        print("Deep research error:", e)
-        answer = "Something went wrong in deep research mode."
+        print(f"Deep research error: {e}")
+        answer = f"Deep research encountered an error. Please try again."
         sources = []
 
     memory.add(ws, "assistant", answer)
@@ -752,13 +757,20 @@ def analyze_mode(req: ModeRequest):
     
     memory.add(ws, "user", q)
     
-    # Run the AnalysisGraph pipeline
-    state = analysis_graph.run(q)
-    
-    answer = state.get("answer", "No analysis generated.")
-    sources = state.get("sources", [])
-    links = state.get("links", [])
-    follow = state.get("followups", [])
+    try:
+        # Run the AnalysisGraph pipeline
+        state = analysis_graph.run(q)
+        
+        answer = state.get("answer", "No analysis generated.")
+        sources = state.get("sources", [])
+        links = state.get("links", [])
+        follow = state.get("followups", [])
+    except Exception as e:
+        print(f"Analysis error: {e}")
+        answer = f"Analysis encountered an error: {str(e)[:100]}"
+        sources = []
+        links = []
+        follow = []
     
     # Get related images
     images = tavily_images_safe(q)
@@ -916,13 +928,20 @@ def web_search_mode(req: ModeRequest):
     
     memory.add(ws, "user", q)
     
-    # Run the WebSearchGraph pipeline
-    state = web_graph.run(q)
-    
-    answer = state.get("answer", "No answer generated.")
-    sources = state.get("sources", [])
-    links = state.get("links", [])
-    follow = state.get("followups", [])
+    try:
+        # Run the WebSearchGraph pipeline
+        state = web_graph.run(q)
+        
+        answer = state.get("answer", "No answer generated.")
+        sources = state.get("sources", [])
+        links = state.get("links", [])
+        follow = state.get("followups", [])
+    except Exception as e:
+        print(f"Web search error: {e}")
+        answer = f"Web search encountered an error: {str(e)[:100]}"
+        sources = []
+        links = []
+        follow = []
     
     # Get images separately
     images = tavily_images_safe(q)
@@ -951,12 +970,23 @@ def rag_mode(req: ModeRequest):
     
     memory.add(ws, "user", q)
     
-    # Run the RAGOnlyGraph pipeline
-    state = rag_graph.run(q, ws)
-    
-    answer = state.get("answer", "No answer generated.")
-    sources = state.get("sources", [])
-    follow = state.get("followups", [])
+    try:
+        if rag_graph is None:
+            # LITE_MODE fallback
+            answer = "RAG mode requires document uploads. In lite mode, please use Web Search instead."
+            sources = []
+            follow = []
+        else:
+            # Run the RAGOnlyGraph pipeline
+            state = rag_graph.run(q, ws)
+            answer = state.get("answer", "No answer generated.")
+            sources = state.get("sources", [])
+            follow = state.get("followups", [])
+    except Exception as e:
+        print(f"RAG error: {e}")
+        answer = f"RAG mode encountered an error: {str(e)[:100]}"
+        sources = []
+        follow = []
     
     memory.add(ws, "assistant", answer)
     
@@ -983,14 +1013,30 @@ def agentic_mode(req: ModeRequest):
     memory.add(ws, "user", q)
     print(f"\n🤖 AGENTIC MODE (LangGraph): {q}")
     
-    # Run the AgenticRAGGraph pipeline
-    state = agentic_graph.run(q, ws)
-    
-    answer = state.get("answer", "No answer generated.")
-    sources = state.get("sources", [])
-    links = state.get("links", [])
-    images = state.get("images", [])
-    follow = state.get("followups", [])
+    try:
+        if agentic_graph is None:
+            # LITE_MODE fallback - use web search
+            state = web_graph.run(q)
+            answer = state.get("answer", "No answer generated.")
+            sources = state.get("sources", [])
+            links = state.get("links", [])
+            images = tavily_images_safe(q)
+            follow = state.get("followups", [])
+        else:
+            # Run the AgenticRAGGraph pipeline
+            state = agentic_graph.run(q, ws)
+            answer = state.get("answer", "No answer generated.")
+            sources = state.get("sources", [])
+            links = state.get("links", [])
+            images = state.get("images", [])
+            follow = state.get("followups", [])
+    except Exception as e:
+        print(f"Agentic error: {e}")
+        answer = f"Agentic mode encountered an error: {str(e)[:100]}"
+        sources = []
+        links = []
+        images = []
+        follow = []
     
     memory.add(ws, "assistant", answer)
     print(f"  ✅ AgenticGraph: Completed with {len(sources)} sources")

config/config.py

@@ -8,8 +8,8 @@ class Config:
     GROQ_API_KEY = os.getenv("GROQ_API_KEY")
     TAVILY_API_KEY = os.getenv("TAVILY_API_KEY")
     
-    # Use lighter model for free tier
-    LLM_MODEL = os.getenv("LLM_MODEL", "openai/gpt-oss-20b")
+    # Use valid Groq model - llama3-8b-8192 is fast and free
+    LLM_MODEL = os.getenv("LLM_MODEL", "llama3-8b-8192")
 
     CHUNK_SIZE = 400
     CHUNK_OVERLAP = 80
@@ -19,19 +19,13 @@ class Config:
 
     @classmethod
     def get_llm(cls):
-        """Return chat LLM instance with tool calling disabled."""
+        """Return chat LLM instance."""
         if not cls.GROQ_API_KEY:
             raise RuntimeError("GROQ_API_KEY missing in .env")
-        llm = ChatGroq(
+        return ChatGroq(
             groq_api_key=cls.GROQ_API_KEY,
             model_name=cls.LLM_MODEL,
             temperature=0.7
         )
-        # Disable tool calling by binding empty tools list
-        try:
-            return llm.bind(tools=[])
-        except:
-            # Fallback if bind doesn't work
-            return llm
 
 

render.yaml

@@ -17,4 +17,4 @@ services:
       - key: LITE_MODE
         value: "true"
       - key: LLM_MODEL
-        value: "openai/gpt-oss-20b"
+        value: "llama3-8b-8192"