Notifications
Clear all
Topic starter
19/07/2026 1:41 am
Hey everyone! I'm just getting started with BabyAGI and trying to set it up for some internal project roadmapping experiments.
I see most guides use OpenAI, but I'd prefer to run a local model for cost and privacy. Has anyone successfully integrated BabyAGI with a custom LLM like Llama 3 or Mistral? What was your setup experience like?
I'm especially curious about configuration changes and performance. Did you run into any hiccups with task generation or execution using these open models? Any recommendations on which model plays nicest with the BabyAGI framework?