Codex Usage details show GPT-6.2 Astra records; OpenAI has not yet commented
BIBIBI
AT A GLANCE
Multiple users said Codex usage details showed gpt-6.2-astra records, but OpenAI has not yet confirmed this.
Article
Codex Usage details show GPT-6.2 Astra, suggesting a limited rollout test
PANews November 1 reports that multiple users posted on X and in developer communities that their Codex usage details showed a "gpt-6.2-astra" model record, while the model selector still displayed GPT-6 Astra. Screenshots from some users showed that approximately 19.8%of requests were assigned to the model, suggesting that OpenAI may be conducting a limited rollout test. Some users also said they did not find related records. As of publication, OpenAI had not issued an official statement, and the model had not appeared in API.
PANews reports that multiple users posted on X and in developer communities that their Codex usage details showed a “gpt-6.2-astra” model record.
02
The report states that the model selector still displays GPT-6 Astra。
03
Screenshots from some users showed that approximately 19.8%of requests were assigned to the model.
04
Some users also said they did not find related records.
05
As of publication, OpenAI had not issued an official statement, and the model had not appeared in API.
AI-assisted interpretation
The following is analysis, separate from reported facts. Verify important claims independently.
Some users saw a new model name in their usage records, but the selectable-model interface still displayed GPT-6 Astra. Because not all users saw the record and OpenAI has provided no explanation, it can currently only be regarded as an unconfirmed sign of testing.
Why it matters to readers
If these records do indeed come from a small-scale test, some Codex requests may be being assigned to a different model; however, the available evidence is insufficient to confirm whether the model formally exists or will be publicly available.
Regular users do not need to adjust their usage based on this for now, and can continue monitoring the model selector,API and OpenAI’s official statements for changes.
Risks and unknowns
It cannot be confirmed whether “gpt-6.2-astra” is an official model, an internal identifier, or a display anomaly.
It cannot be confirmed whether the approximate 19.8%allocation ratio applies to anyone other than the user in the screenshot.
It cannot be confirmed what the scope or duration of the test is, or whether it will be publicly launched.
The claims mainly come from user posts and screenshots and lack official confirmation from OpenAI.
Related Developments
Loading event timeline…
Related concepts
Usage details
This term is not in the glossary yet. Browse related concepts in the glossary.
