I noticed something unexpected in my ChatGPT usage statistics and I’d like to understand whether this is expected behavior or a reporting/routing issue.
In the last 7 days, my activity page shows 123 turns. When I switch the chart to “By model”, a substantial portion of the activity—especially on September 6 and 7—is attributed to gpt-5.6-luna. The same chart also shows usage of gpt-5.6-sol and gpt-5.6-astra.
The important point is that I have never intentionally selected or used GPT-5.6 Luna.
I then checked the same activity using “By surface”. All 123 turns are attributed to Desktop.
So, as far as I can tell:
I did not knowingly select Luna.
A significant number of my turns are nevertheless reported as gpt-5.6-luna.
All of this activity is reported under the Desktop surface.
The Luna usage increases sharply on September 6–7.
I’m trying to understand what exactly these statistics represent.
Does ChatGPT sometimes route Desktop conversations to Luna automatically, even when another model is selected or expected? Could Luna be used internally as a fallback or for certain requests without being shown in the model selector? Or is the “By model” activity chart measuring something different from the model that actually generated the visible response?
I’m particularly interested in knowing whether other users are seeing the same thing.
I’m attaching screenshots of both By model and By surface.
If this is expected behavior, an explanation of how model routing is reflected in the activity statistics would be very helpful. If it is not expected, this may be useful as a reproducible reporting/routing issue.
Thanks for the info, nice to get such questions out for others to see.
The following is a standard reply template I use, do not read anything more into it other than useful information.
You are welcome to discuss Codex issues here. However, the official place to report and track Codex bugs is the OpenAI Codex GitHub issue tracker.
I asked ChatGPT to search for the closest related issue, and it identified:
If that issue describes the same problem you are experiencing, consider adding a reaction to the GitHub issue itself, rather than only reacting to this forum topic. GitHub reactions can help the maintainers gauge how many users are affected.
If it does not match your problem closely enough, search the existing Codex issues for a better match. If you cannot find one, consider opening a new issue and include as much relevant information as possible, such as your Codex version, operating system, environment, reproduction steps, expected behavior, actual behavior, and any relevant logs or error messages.
reaction to the GitHub issue itself, rather than only reacting to this forum topic. GitHub reactions can help the maintainers gauge how many users are affected.