1. Verbose tool output
Every shell command, build run, and test invocation sends its full output back to the model. A passing test suite can dump hundreds of lines of dots, timing data, and progress bars. A failing build can produce thousands of lines of stack traces, deprecation warnings, and dependency resolution noise. Codex reads all of it, every token counts against your usage, and the actual signal is usually buried in 5% of the output.
2. Repeated context across turns
Codex re-sends conversation history with every turn. The same file content, the same tool definitions, the same earlier responses get replayed for the model on each step. In a long debugging session this compounds quickly: the tenth message in a thread can carry ten times the context of the first.
3. Multi-step debugging sessions
Real debugging is rarely one prompt. You run a test, read the failure, ask Codex to investigate, it reads three files, runs another command, reads more files, suggests a fix, you run the test again. Each step adds more tool output, more file content, more context. A 30-minute debugging session can easily consume more of your usage window than a full afternoon of writing code.
4. Large codebase reads
Asking Codex to "explore the codebase" or "find where X is defined" pulls in big chunks of source files, often with surrounding context Codex doesn't strictly need. If your repo has long generated files, JSON fixtures, or HTML templates, a single search can move tens of thousands of tokens through the model.
The Codex twist: usage is now metered by tokens
In April 2026, OpenAI moved Codex on Plus, Pro, and Business plans away from simple message counts to token-based metering, in line with API usage. That makes the patterns above bite harder: the size of every request now maps directly to how fast you drain your 5-hour and weekly windows, so a few noisy, token-heavy turns cost far more than they used to. And when you do hit the cap, your built-in options are to buy extra credits or drop to a smaller model — paying with either money or quality. Cutting the tokens in each request sidesteps that trade entirely.
How to fix it
The fix is the same in every case: remove low-signal content before Codex sees it, while preserving the structure and information the model actually needs. That is what Headroom does — it sits between Codex and the API and reversibly compresses repetitive logs and verbose documents before forwarding them, so the model can still pull back anything it needs. Token spend drops by ~50%, with no measurable drop in output quality.
For a deeper walk-through of the tools that handle each source of waste, see our Codex cost guide. If your problem is hitting the 5-hour plan limit rather than billing, our Codex usage limits guide covers that specifically.