CN:
我最近在 Claude Code 中测试 gpt-5.4，但它的响应和完成任务速度非常慢。
我想请教一下已经使用过它的朋友：
gpt-5.4在 Claude Code 中通常会比 Claude 模型更慢吗？- 这个延迟是由模型本身、API 服务商，还是 Claude Code 引起的？
- 有没有推荐的设置可以提高速度？
- 目前在 Claude Code 中，哪个模型最适合用于编程任务？
gpt-5.4适合用于大型代码修改吗？还是更适合用于规划和代码审查？
我的设置：
- 工具：Claude Code
- 模型：
gpt-5.4 - 问题：响应非常慢 / 任务完成时间很长 / 使用了很多 skills
如果有朋友测试过，希望能分享一些建议。谢谢！
EN:
I have been testing gpt-5.4 with Claude Code, but it is taking a very long time to respond and complete tasks.
I wanted to ask people who have already used it:
- Is
gpt-5.4normally slower in Claude Code compared to Claude models? - Is the delay caused by the model itself, the API provider, or Claude Code?
- Are there any recommended settings to improve speed?
- Which model works best with Claude Code for coding tasks right now?
- Should I use
gpt-5.4for large code edits, or is it better for planning/review only?
My setup:
- Tool: Claude Code
- Model:
gpt-5.4 - Issue: Very slow responses / long task completion time / using a lot of skills
Any advice from people who have tested this would be helpful.
6 个帖子 - 5 位参与者
来源: linux.do查看原文