fix: Do a better job of mapping Langchain to LiteLLM (#5233)

src/backend/base/langflow/base/agents/crewai/crew.py

@@ -1,6 +1,7 @@
 from collections.abc import Callable
 from typing import Any, cast
 
+import litellm
 from crewai import LLM, Agent, Crew, Process, Task
 from crewai.task import TaskOutput
 from crewai.tools.base_tool import Tool
@@ -70,9 +71,12 @@ def convert_llm(llm: Any, excluded_keys=None) -> LLM:
         msg = "Could not find model name in the LLM object"
         raise ValueError(msg)
 
-    # Normalize Ollama with prefix TODO: Handle all litellm supported models
-    if llm.dict().get("_type") == "chat-ollama":
-        model_name = f"ollama/{model_name}"
+    # Normalize to the LLM model name
+    # Remove langchain_ prefix if present
+    provider = llm.get_lc_namespace()[0]
+    if provider.startswith("langchain_"):
+        provider = provider[10:]
+        model_name = f"{provider}/{model_name}"
 
     # Retrieve the API Key from the LLM
     if excluded_keys is None:
@@ -188,8 +192,13 @@ class BaseCrewComponent(Component):
         return step_callback
 
     async def build_output(self) -> Message:
-        crew = self.build_crew()
-        result = await crew.kickoff_async()
-        message = Message(text=result.raw, sender=MESSAGE_SENDER_AI)
+        try:
+            crew = self.build_crew()
+            result = await crew.kickoff_async()
+            message = Message(text=result.raw, sender=MESSAGE_SENDER_AI)
+        except litellm.exceptions.BadRequestError as e:
+            raise ValueError(e) from e
+
         self.status = message
+
         return message