Codex should adapt output wrapping to the actual terminal or input area width
What variant of Codex are you using?
CLI
What feature would you like to see?
Summary
Codex output does not appear to properly adapt to the actual width of the current terminal, input box, or output area.
When using Codex in a wide terminal window, especially one with a large X-axis and relatively small Y-axis, the output still seems to be formatted as if the available width were much narrower. This results in very short lines and excessive line breaks.
This may not be limited to Codex itself. It may be related to the API's default output preference or formatting behavior, because a similar issue can also be observed when using the API through Cherry Studio.
Problem
In wide-layout environments, the model often produces responses with unnecessarily frequent wrapping. Even when there is plenty of horizontal space available, the output is still split into many short lines.
This significantly reduces readability. Because so much vertical space is consumed by short wrapped lines, the amount of visible context that a human can read at once becomes much smaller than it should be.
Example scenarios
- Use Codex in a wide terminal window, where the X-axis is large and the Y-axis is relatively small.
- Ask Codex to produce a normal explanatory response, code-related analysis, or multi-step debugging guidance.
- Observe that the output still uses short lines and frequent line breaks instead of taking advantage of the available horizontal width.
A similar behavior can also be seen when using the API through Cherry Studio, which suggests that this may be related to the model/API's preferred formatting style rather than only a Codex terminal rendering issue.
Expected behavior
Codex, or the API output style behind it, should better adapt to the actual available width of the terminal, input box, or output area.
In a wide terminal or wide-screen environment, the output should make better use of the horizontal space so that each line contains more useful content and the user can read more context at once.
Actual behavior
The output appears to use a fixed or overly conservative wrapping style. As a result, the response contains many short lines and excessive line breaks, even when the display area is wide enough to support longer lines.
Why this matters
For users working in wide terminals, wide-screen coding environments, or API-based desktop clients, the current formatting reduces the amount of visible context and makes responses harder to scan.
This is especially noticeable when reading explanations, debugging steps, command output analysis, or multi-step reasoning.
Suggested improvement
Codex and/or the API output behavior should dynamically consider the available rendering width and adjust the text wrapping strategy accordingly.
Ideally, the model should avoid overly short lines when the available display width is large, so that wide terminals and wide-screen clients can provide a better reading experience.
Additional information
_No response_