diff --git a/src/marvin/bots/base.py b/src/marvin/bots/base.py
index c81c6d96f..57f932ace 100644
--- a/src/marvin/bots/base.py
+++ b/src/marvin/bots/base.py
@@ -5,6 +5,7 @@
 from typing import TYPE_CHECKING, Any, Callable
 
 import pendulum
+from openai.error import InvalidRequestError
 from pydantic import Field, validator
 
 import marvin
@@ -490,9 +491,20 @@ async def _call_llm(self, messages: list[Message]) -> str:
         if marvin.settings.verbose:
             messages_repr = "\n".join(repr(m) for m in langchain_messages)
             self.logger.debug(f"Sending messages to LLM: {messages_repr}")
-        result = await self.llm.agenerate(
-            messages=[langchain_messages], stop=["Plugin output:", "Plugin Output:"]
-        )
+        try:
+            result = await self.llm.agenerate(
+                messages=[langchain_messages], stop=["Plugin output:", "Plugin Output:"]
+            )
+        except InvalidRequestError as exc:
+            if "does not exist" in str(exc):
+                raise ValueError(
+                    "Please check your `openai_model_name` and that your OpenAI account"
+                    " has access to this model. You can select an OpenAI model by"
+                    " setting the `MARVIN_OPENAI_MODEL_NAME` env var."
+                    " Read more about settings in the docs: https://www.askmarvin.ai/guide/introduction/configuration/#settings"  # noqa: E501
+                )
+            raise exc
+
         return result.generations[0][0].text
 
     async def interactive_chat(self, first_message: str = None):