Testing Moonshot AI's Kimi K3 Inside Claude Code
philippdubach.com
philippdubach.com
A test of how different models perform in more of a model-agnostic harness like OpenCode would be much more interesting, perhaps paired with a control experiment of how those same models performed on the same task when using their respective native harnesses.
A lot of folks don't seem to understand that it's the harness and its behavior paired with a particular model that produces the result, and you have to focus on both aspects. He barely acknowledges the importance of the harness in his write-up, beyond the two notes above.
The pass@1 scores are 50/45/40 for OpenCode/Cursor/Claude Code respectively.
https://artificialanalysis.ai/agents/coding-agents#harness-c...
I've seen other benchmarks where Pi also outperforms Claude Code both in model performance and in much reduced token usage.