Study: coding agent framework complexity doesn't broadly raise pass rates
What happened
On October 24, The Decoder reported on Liu and Han's study of coding agent frameworks. It compared Claude Code, OpenCode and mini-SWE-agent and found that differences in framework complexity did not broadly raise task pass rates, while per-task cost with the same model varied by as much as three times. The report gives no specific pass rates or per-task cost figures for each framework.
Written by AI from the coverage · updated 3 hours ago
Coverage
Follow the reports to see the story from different sides.
- The DecoderFancy harnesses make coding agents more expensive but not better, study finds
Liu 与 Han 的研究发现，Claude Code、OpenCode 和 mini-SWE-agent 的复杂度差异未普遍提高任务通过率，同一模型的单任务成本却可相差三倍。
Heat over time
Heat now 9·Comparable peak 10（Oct 24 19:00）·Comparable change over 24 hours –
The trend only compares accounts observed without gaps, so its range may be smaller than the current heat. Hover or tap the chart for each hour; the left and right arrow keys step through it.