I second qwen. It is very useable model. Sonnet is of course better (also 200k context vs 32k), but sometimes I just cannot take the risk of letting any sensiti

I second qwen. It is very useable model. Sonnet is of course better (also 200k context vs 32k), but sometimes I just cannot take the risk of letting any sensitive data "escape" in the context so i use qwen and it is pretty good.