Migrating from Claude Code to Codex — motivation and day one

I’ve been using Claude Code for a long time. It gradually became my main development tool, surrounded by skills, plugins, memories, and CLAUDE.md files across my projects. Everything works, so I decided to change it 🙃

For the next month, I’ll work mainly with OpenAI Codex. At first, both tools will run side by side. Will they get in each other’s way? I don’t know. Will I make a mess across my repositories? Quite possibly. But I have to start somewhere.

The first obvious question — why?

Several reasons piled up: curiosity, the policy around third-party tools, increasingly positive feedback, and price. None would probably be enough on its own. Together, they were enough for me to give Codex a month to prove itself.

I want to learn a new tool

There is no grand philosophy here. Codex is developing quickly, and since I love innovation and new things, I want to test it on my own projects.

Reading yet another “Claude Code vs Codex” comparison won’t tell me how Codex handles my repositories, rules, and workflow. A month of using it will.

Better policy

OpenAI takes a much better approach to third-party tools. One example? You can use OpenClaw with ChatGPT/Codex subscription authentication, without paying separately for every API call.

Claude chose a different path here. So did I.

Increasingly positive feedback

Not that long ago, I wasn’t considering Codex at all. More than that, I had become heavily biased: I ignored OpenAI’s product and strongly favored Anthropic. Claude worked well, so every new piece of information about Codex was easy to dismiss with a shrug.

Eventually, there were too many positive reports about speed, code quality, and limits to keep pretending I couldn’t see them.

Price

Codex is cheaper, especially when the reference point is Claude Max 5x, which I would need at my level of usage. Of course, a lower price is useless if the tool produces worse code. After the first few days, I haven’t seen that problem.

First day, first observations

Reasons are one thing, but working with the tool is what ultimately matters. My first days with Codex brought several surprises — mostly positive, which was mildly inconvenient for someone who had been successfully ignoring OpenAI.

A free month

I used to have a paid OpenAI subscription, but later moved entirely to Claude. After a long break, I returned and received a small gift: one free month of ChatGPT Plus.

At first, I barely used it. My attachment to Claude kept winning, while Plus waited patiently in the background. Over the last few days, however, Anthropic’s limits started disappearing noticeably faster. Fable burned through my weekly limit halfway into the week. By Wednesday, I was out of tokens — and that was when OpenAI started filling the gap.

The free month of Plus saved the situation. Today, I bought a paid Codex subscription and cancelled Claude Max 5x. I still have four days left, but it won’t renew.

It is faster

Codex and ChatGPT generate answers and code noticeably faster. I didn’t measure it with a stopwatch because I have more interesting things to do, but I can feel the difference on practically every task. Less time watching the generation animation, more time working. Good deal.

The limits disappear more slowly

With a similar workflow, Codex limits drain much more slowly than Claude Code’s. After several larger tasks, I can still work instead of analyzing the usage bar like a stock chart.

We’ll see whether that remains true for the whole month. The first few days look very promising.

Migration was suspiciously easy

I expected to move the configuration manually. There was no need. OpenAI added a graphical “Import to Codex” flow to the app, with support for Claude Code configuration. That is what I used — no commands and no manually moving files around.

The migration covered skills, memories, plugins, and the rest of my configuration. A few clicks later, most of my existing environment was ready. Almost boring.

There was one problem. Shared repositories did not get equivalents of their CLAUDE.md files. One prompt later, Codex analyzed the projects and prepared AGENTS.md files for them. Problem solved.

I don’t have to relearn how to work

I mainly use Codex CLI. Moving from Claude Code required no change in habits: I open a repo, describe the task, and the agent analyzes files, makes changes, and runs tests.

The familiar features are there too. Sessions can be resumed with resume. There is also an equivalent of --dangerously-skip-permissions: --dangerously-bypass-approvals-and-sandbox, for people who enjoy long flags and living on the edge. Thankfully, there is also a --yolo alias.

The entry barrier? Practically nonexistent.

Code quality is comparable

This was the most important point for me. Faster answers and higher limits mean nothing if I have to fix half of the generated code afterward.

So far, the quality is comparable to Claude Code. Well-prepared AGENTS.md files were enough for Codex to understand the project structure, commands, and my expectations. The first changes landed at the same level as before.

The first score

If I’m comparing two tools, there has to be a table. Otherwise, the internet might refuse to recognize the comparison.

Area	Claude Code	Codex	Result
Speed	Noticeably slower	Answers and code appear faster	Codex
Limits	A weekly limit could disappear halfway through the week	They drain more slowly with similar usage	Codex
Price	Max 5x would cost more at my usage level	Cheaper subscription	Codex
Code quality	High and proven over a long period	Still testing	In progress
Policy	More restrictive approach	OpenClaw works with a ChatGPT/Codex subscription	Codex

After the first few days: Codex 4, one result still in progress. Claude has four days of active subscription left to protest.

What next?

I bought a month-long subscription, and I’ll stay with Codex during that time. For the first few days, I’ll continue using both tools so I can compare them on the same types of tasks.

I’ll share the next observations around the middle of the subscription. The final part will come at the end of the month — by then, we’ll know whether this was a migration or just a longer affair with a new tool 😉

Stay tuned!