I am getting Failed to start execution / Failed to run automation (500) in CrewAI Studio.
I tested the issue with a minimal setup: one agent, one simple task, no external tools, and an explicitly assigned model Meta-Llama-3.3-70B-Instruct.
The LLM connections appear active, but the execution still fails immediately before any trace is created.
Please check the project/account execution backend and the LLM connection on your side.