Skip to main contentTailor AI LogoTailor AI
    Engineering Blog
    Agentic Coding1 min read

    I've been running Codex and Claude Code side by side all day. Each has one major problem.

    Greg Bayer
    Greg Bayer

    I've been running out of Claude Code tokens on the $200 plan, so I decided to dive into Codex for active development. I've been running 5+ parallel coding tasks pretty much all day. I also make heavy use of each harness's Chrome extension for web development loops, where the AI can review the results of its own changes and iterate.

    Here are my conclusions:

    CC and Codex are both reliable and pretty similar in quality, except each has one major problem the other doesn't.

    CC issue: Claude-in-Chrome seems to have a serious memory leak. It constantly crashes Chrome, and sometimes my whole machine, if I don't kill it fast enough. I'm also running out of tokens every Thursday now, even without using Fable.

    Codex issue: It's horrible at frontend design. In this area, Claude Code feels like a senior developer with an eye for design, while Codex feels like an intern with basically no design sense at all.