Post

@CommandCodeAI Okay I've to admit either the model is just too too good or all these people are just using it for things apart for coding. Because I've used it and while it's definitely better than other open source models but still not at the same level of Opus(obviously before nerfing)
English

@CommandCodeAI Great!
I analyzed the cost of GLM 5.2 vs Frontier LLMs, in different use-case scenarios.
Check it here:
shayanspiel.github.io/GLM5.2-Cost-Br…
English

@CommandCodeAI hey can you guys make it possible to use local models with your harness and if its possible how do i do it and thank you
English

@CommandCodeAI when my go usage rans out can i buy another dollar and get same 10x value (please say yes)
English

@CommandCodeAI Thank you, since infrastructure upgrade in your backend, now commandcode performance is better and much faster. I see the difference now, from up to 5 minutes in thinking for some task, now it can take 2 minutes. I use Deepseek Pro and Minimax btw.
English

@CommandCodeAI My command code subscription got exchausted in single day on complex task 😔 Cannot use it. Switched to Opencode
English

@CommandCodeAI when quality gets close enough, price becomes the dominant factor in developer tool usage
that's the pattern showing up here
English

@CommandCodeAI It does not appear for me. And also didn´t find in the website model list.
English

@CommandCodeAI Is it correct that Deepseek pro is cheaper at 40x?
English

@CommandCodeAI Your tool is available as vscode agentic chat as well just like copilot?
English

@CommandCodeAI Please give me 1 billion token 🤣
English

@CommandCodeAI Could you please make it easier by directly sharing how many millions of tokens a user gets with a specific pricing plan? I am tired of the "x times" marketing terminology.
English






