Notifications
Clear all
Topic starter
20/07/2026 5:11 am
Hey everyone! I'm trying to get more hands-on with my AI coding setup and want to move away from the cloud APIs for some projects. I've been using Aider with GPT-4, but I'd love to run everything through a local or private endpoint for better control and cost management.
What would you recommend for a setup? I'm especially curious about the config steps to point Aider at a private LLM, like a local Ollama or an OpenRouter endpoint. Any gotchas with the model's capabilities or the prompt formatting I should watch out for? 😊