Comment by verdverm

This, I've switched to qwen36moe on a spark and it's on par with gemini-3-flash. It's way better than I expected and in another ~6 months I expect things to be even better for open models of this size.

My long term expectation is that the big labs will build big models primarily for training and distilling to economically viable models. Even if the model capabilities don't plateau so soon, I think the economics of this will.