GPT‑5.1 Codex Max uses inconsistent tools for file modifications

What version of Codex is running?

0.61.0

What subscription do you have?

Pro

Which model were you using?

_No response_

What platform is your computer?

_No response_

What issue are you seeing?

When using the GPT-5.1 Codex Max model, I notice that its file-editing tool calls are inconsistent.
Sometimes the model uses the built‑in edit, add, or delete tools to modify files; at other times it invokes the apply_patch tool instead. Both approaches ultimately modify the files correctly, but this inconsistent behaviour feels random and makes the model’s tool usage unpredictable. From a UI perspective, the built‑in edit/add/delete tools provide clearer feedback to the user than apply_patch.

What steps can reproduce the bug?

1. Use Codex 0.61.0 and select the gpt‑5.1-codex‑max model.
2. In a conversation, request the model to make changes to a file (e.g., ask it to modify some code, add a new function, or delete a line).
3. Observe which tool call is made. In some cases it will call the built-in edit/add/delete tools, and in other cases it will call apply_patch. There is no obvious pattern; it seems random.

What is the expected behavior?

The model should consistently use the same mechanism for file modifications. Ideally, it should call the built‑in edit, add, or delete tools for file operations, rather than switching between those and apply_patch. Aligning the tool calls would make the model’s behaviour more predictable and improve the UI/UX.

Additional information

This inconsistency is not blocking my workflow but is confusing, and I believe aligning tool usage would provide a better experience.