I was very skeptical about the whole concept of local models, and i cannot explain how wrong i was.
Let me share this very nice money saver for small tasks.
LMStudio acts as your repository for local models. Continue is the opensource agentic workforce that allows you to replicate most of the copilot features when you combine the local models with your agentic code builder it's basically copilot for free.
The only limits are the resources you can allocate, and in my case are not much...
And the next step... is to leverage ollama to build your own batch llm service and ditch api call costs.