Generated from the live deployment on 2026-09-08T01:05:38+00:00. Do not edit by hand.
These are the models the portal can actually use. If a model is not in this table, it is not available to anyone, whatever another guide may say.
| Name you pick | Model | Deployment type | Version policy |
|---|---|---|---|
gpt-5.4 |
gpt-5.4 2026-03-05 | DataZoneStandard | NoAutoUpgrade |
gpt-5.4-mini |
gpt-5.4-mini 2026-03-17 | DataZoneStandard | NoAutoUpgrade |
gpt-5.6-sol |
gpt-5.6-sol 2026-07-09 | DataZoneStandard | NoAutoUpgrade |
gpt-5.2 |
gpt-5.2 2025-12-11 | DataZoneStandard | NoAutoUpgrade |
gpt-5.3-codex |
gpt-5.3-codex 2026-02-24 | DataZoneStandard | NoAutoUpgrade |
model-router |
model-router 2025-11-18 | DataZoneStandard | NoAutoUpgrade |
text-embedding-3-large |
text-embedding-3-large 1 | DataZoneStandard | NoAutoUpgrade |
Every model is DataZoneStandard. That
is the setting that decides where your words are processed: the APAC
data zone, never worldwide. Choosing a different model from this list
does not change where your work goes, and does not change what records
you are allowed to see.
Which one should I use?
| If you are | Use | Why |
|---|---|---|
| Louise or Yuki, doing customer service | gpt-5.4-mini |
Fast, strong in Traditional Chinese and Japanese, and much cheaper per request |
| Editing, or executive and admin work | gpt-5.4 |
Quality shows most in editorial judgement |
| Writing or reviewing code | gpt-5.3-codex |
gpt-5.6-sol and model-router are available
but are nobody's default: their cost cannot be forecast, so they should
not carry routine work.