qwen3-coder-next is really good! all you need is 50GB vram, i pray for the day that we can all run our coding agents locally
There’s currently no open source LLM capable of handling decent coding tasks ? :o