Hello. I'm a beginner AI programmer.
Originally, my company gave me a Claude Code account so I was using the best model, but..
Since I'm about to leave the company soon, I was looking for an AI model to work alone from home..
I heard there were people using local AI, so I asked in the community below,
and many people said qwen3.7-27b was good, so I installed it on my Mac.
At first, I heard they use something called LM Studio, so I installed LM Studio and set it up this way and that way,
but when I searched the internet, most of them use Agent based on Visual Studio Code or
something like that, and it was a bit insufficient for me..
I figured I should use the Copilot AI Agent that comes with Visual Studio 2026.. so I can better use the
complex features of VS 2026..
I discovered something called Ollama that I could access, and while searching for what Ollama actually is,
as a result of setting it up this and that way.. I successfully accessed my Mac in AI Agent mode in VS 2026's Copilot Chat!
And as shown below, I just threw a project called CoreService that I wrote at it to analyze it..
My MacBook Pro worked hard with a whirring sound~ and the results came out so well ㄷㄷㄷ
It's shocking. To this extent.. I'm not sure about the previous Opus, but it seems to work up to that earlier version..
I need to test it thoroughly.. the LLM world is amazing..
Is it Kimi 2.7? I wonder if these large models are even better.. ㄷㄷㄷ
