And you let local QWEN write the code for you? Is the output any good or comparable to frontier models?
And you let local QWEN write the code for you? Is the output any good or comparable to frontier models?